US2434922A - Pulse amplitude selector system - Google Patents

Description

Jan. 27, 194s.

D. DQ GRIEG PULSE AMPLITUDE SELECTOR SYSTEM 5 Sheets-'Sheet 2 All /P la ze 7/-13 23a /ZZL F22 I 24 28d @l J4 "J5 INVENTOR.

A T WPA/EY Jan. 27,1948. D, D, GREG v K 2,434,922

PULSE AMPLITUDE SELECTOR SYSTEM Filed Nov. 2, 1944 5 Sheets-Sheet 5 I '0. A dg. 4. 16'

P01555 0F CURVE' i dpr/6.6.)

IN VN TUR.

DONALD D. GIP/R? Jan. 27, 194s. D. D. GREG A2,434,922

PULSE AMPLITUDE SELECTOR SYSTED Filed Nov. 2, 1944 5 Sheets-Sheet 4 /NPUT D L f1/M Law -wbl

0 v 45 il v i l l l l l l l I p l l l I v Twig-ff 1| m 7/ I [K f`\ INVENTOR. mma a. 6mm

Amm

Jan. 27, 1948. 0, D, GRIEG. 2,434,922

PULSE AMPLITUDE SELECTOR SYSTEM Filed Nov. 2, 1944 5 Sheets-Sheet 5 IN VEN TOR. 00A/4L 0 0. GIP/f6 Patented Jan. 27, 1948 PULSE AMPLITUDE BELECTOB SYSTEM Donald D.',Grieg, Forest Hills. N. Y.. assigner to Federal Telephone and Radio Corporation, New York, N. Y., a corporation of Delaware Application November 2, 1944, Serial No. 561.554

23 Claims. (Cl. 250-27) This invention relates to radio pulse systems and more particularly to systems for discriminating between pulses in accordance with the amplitude of the pulses. This application is a continuation in part of my copending application Serial No. 467,516, now abandoned, illed December 1, 1942.

Excluding the condition of superposition of pulses, there exist three possibilities of relative amplitude variation during interference between series of wanted and unwanted electrical pulses substantially the same time duration on width. These possibilities are where the amplitudegof the unwanted pulse is smaller, equal to, or. larger than the amplitude of a wanted pulse. Where the unwanted pulse is smaller, it may be eliminated by simple clipping. The probability of two different sources of pulses having the same amplitude and thus causing interference at a receiver is so small as to be negligible. However, the elimination of unwanted pulses of greater amplitude than wanted pulses has heretofore presented a more difficult problem.

It is an object ofthis invention to discriminate selectively between pulses dill'ering in amplitude so as to eliminate unwanted pulses regardless of the magnitude of their amplitude relative to wanted pulses.

The method of my invention may, for example, comprise distorting the leading or trailing,v or both edges of the unwanted pulses of amplitude greater than the amplitude of the wanted pulses to reduce the slope. Such distortion varies the width at least in some part of an unwanted pulse so that by width discrimination of a part of the distorted pulse portion taken between predeterminedamplitude levels and the corresponding pulse portion of the wanted pulse, pulse energy corresponding to the wanted pulse only can be obtained.

In place of the gate. clipping and width discrimination steps, the unwanted distorted pulses and the wanted pulses may be otherwise processed, such as by a differentiating and clipping .operation Differentiation of the pulses provides derivation pulses of one polarity corresponding to the leading edges of the input pulses, and derivation pulses ofopposite polarity corresponding to the trailing edges of the input pulses. The derivation pulses corresponding to the distorted edges will have, after one o r more differentiating operations, smaller amplitudes than the other derivation pulses. The derivation pulses resulting from the differentiating step are clipped at a level at least as high as the amplitude of the pulses derived from the distorted edges,thereby obtaining a pulse output in accordance with the occurrence of the wanted input pulses.

The distortion of leading and/or trailing edges of the larger amplitude pulses may be accomplished by producing dissimilar pulse potentials in response to the larger amplitude pulses which are combined or applied in one way or another to .the larger amplitude pulse energy so as to distort either its leading or trailing edge, or both, thereby distinguishing the unwanted pulses from the wanted pulses by a difference in one or more shape characteristics other than amplitude. When the pulses are differentiated, the distorted edge or edges will not produce pulses increasing and decreasing sharply in amplitude. Instead, the distorted edges of the input pulse produce upon differentiation blunt or curved pulses which. in many cases, may be oi' smaller amplitude than the pulses produced by differentiation of the wanted pulses. Should the blunt pulses thus produced by the distorted edges be oi' greater amplitude than the pulses produced from the edges of the wanted pulse, the differentiating operation may be repeated and, since the blunt pulses corresponding to the edge oi the distorted pulse become less and less sharp, they can be eliminated or at least reduced to such an amplitude that a clipping operation will completely eliminate the eilect of the larger input pulses.

The dissimilar pulse potentials used for distorting the unwanted pulses may be' produced by different methods. For example, each larger input pulse may be clipped and the clipped portion distorted by condenser action, or it may be displaced in time relative to the input pulse, or both. The dissimilar pulse potential and the input pulse may be combined either by an algebraic mixing action (addition or subtraction), or by altering, by means of the dissimilar pulse potential, the gain of an amplifier to which the input pulses are applied.

For a better understanding of the methods of which the selective discrimination between the pulses in accordance with amplitude may be effected;

Figs. 2, 3 and .4 are graphical illustrations showing the steps of operation by which pulses of intermediate amplitude can be selectively removed from a train having pulses with amplitudes less than, equal to, and greater than the wanted intermediate pulses:

Fig. 6 is a schematic wiring diagram oi a pulse width discriminating circuit;

Fig. 6 is schematic illustration of another system by which the method oi this invention may be practiced;

Fig. 7 is a graphical illustration useful in explaining the operation of the system oi Fig. 6.

Fig. 8 is a schematic illustration of a further system of the invention; and

Fig. 9 is a graphical illustration useful in explaining the operation ol the system of Fig. 8.

Referring to Figs. 1, 2 and 3 of the drawings.

a' puise modulated wave or pulse train I is applied to the-input 2 of the system of Fig. 1. The train I is shown in Fig. 3, curve a, to contain pulses A. B and C wherein the pulse B is the wanted pulse which may be modulated in either time or width with intelligence. The pulses A and C which are larger and smaller in amplitude, respectively, than pulse B may comprise two additional channels oi' intelligence or merely represent dummy or interfering pulses.

Since the most diilicult discrimination exists where the wanted and unwanted pulses are of substantially the same width and have leading and trailing edges of substantial identical slope, I have chosen toillustrate my invention with pulses of that characteristic.

The pulses C of less amplitude are eliminated by first clipping the train of pulses along the amplitude level represented by line 3 by passing the wave through a clipper 4. The output'of the clipper 4, curve b, is applied to a tube circuit having a vacuum tube 5 which is biased in accordance with the dynamic curve 6' shown in Fig. 2. whereby only the upper portion Ai of the larger amplitude pulses, which extends beyond level 'L appears in the anode circuit 3 of the tube. Connected between the anode circuit l and ground is a condenser 3 adapted to be charged upon the occurrence oi plate current whereby the pulse portion Ai will be distorted. This distortion provides an output pulse I having curved leading and trailing edges II and I2 substantially as shown in Fig. 2, and curve c, Fig. 3. Energy of the pulses A and B is shunted by a connection I3 about the tube and inverter I4 to the mixer stage I5 which may be a conventional double triode stage with plates in parallel. The pulse III combines in the plate circuit of the mixer stage with the larger amplitude pulse from which it is produced. This combining of the pulse I0 with the pulse A results. where both are ofthe same polarity, in a pulse I8 the leading and trailing edges of which are distorted as indicated by curve d. The lower leading edge portion I1 of the pulse I6 is substantially vertical but the upper leading edge portion I8 curves from the vertical to the trailing edge portion I3, in accordance with the curved leadingedge oi pulse IIl, thereby eliminating the sharp upper left hand corner substantially as indicated by the overlapping pulse shapes of curve c. The upper trailing edge portion I3, curve d, is substantially vertical while the lower trailing edge portion is curved according tc the trailing edge of pulse I0. Thus.

ing to the curvetures oi the portions Il and 2li, respectively.

The pulse output at 23 (Fig. 1) may be applied to a known diilerentiating device 2i whereby alternate positive and negative pulse derivations 2i, 22 and 23, 24. curve e, are produced corresponding in time with the leading and trailing edges of the pulse I3 and the pulse B, respectively. The vertical portion I1 results in a vertical rise 2Ia and curved portion I8 results in an exponential drop substantially as indicated at 2| b. The leading and trailing edges 22a and 22h resulting from the trailing edge portions I9 and 20 are similarly shaped in a negative direction. Differentiation oi' the pulse B results in sharper positive and negative pulses 23 and 24 than the pulses 2| and 22, the pulses 23 and 24 corresponding, respectively to the leading and trailing edges of the pulse B.

Should it be desired to obtain a pulse output in accordance with the leading edges of the puise B, the positive pulses 2| and 23 may be threshold clipped at 21 as indicated by the line 21a and the peak portions 23 and 28 resulting therefrom, curve f. diierentiated. The derivations of such a differentiation oi pulse 28 are a sharp positive pulse 30, curve g, corresponding to the vertical `edge 23a, and a blunt negative pulse 3I corresponding to the sloping edge 28h. Diierentiation of pulse 29 results in positive and negative pulses 32 and 33 both o! which correspond substantially in time to the leading edge of puise B. The pulse 3i, curve g, resulting from the sloping edge 28h, curve f, is of smaller amplitude than pulse 33. By clipping the negative pulses 33 at a level 34 beyond the pulse 3|, a unidirectional pulse 35, curve h, is obtained corresponding substantially to the occurrence of the leading edge of the pulse B.

Should one clipping and differentiation be in sufficient to reduce the pulse energy resulting from portion i8 or 20, curve d, to less amplitude vthan the pulse energy resulting from the leading or trailing edge. oi pulse B, additional difierentiating and clipping operations may be made as indicated at 38 and 31. This would be necessary if the clipping level were chosen at 33 instead of at 34, curve g.

If a unidirectional pulse is desired correspondlng to the trailing edge of the pulse B, the negative pulses 22 and 24, curve e. are threshold clipped as indicated by the clipping level 39, curve e. The output of this clipping operation is diii'erentiated in the same manner as described in connectionwith the pulses of curce f and if necessary the clipping and differentiating operations may be repeated should the slope of the edge 22h be too sharp.

While I have described above a differentiating and clipping operation. such as set forth in my copending application Serial No. 561,553, led oi' even date herewith. for discriminating between distorted unwanted pulses and wanted pulses, it

will be clear that other processing means may be employed once the unwanted pulse has been distorted.

As hereinbefore mentioned. gate clipping and width discriminating operations may be employed. By moving contact 25 (Fig. 1) to connection 4Il pulse energy such as shown in curve d, Fig. 3 and curve i, Fig. 4, is then applied to clipper gate 4I. The gate 4| may be of any known clipper arrangement wherein the pulse energy is clipped between two gate limits such as indicated at 42 and 43 on curve i. The portions clipped between limits 42 and '43 may or may not be amplified depending upon the clipper circuit. Preferably `the pulse portions are amplified and then applied as negative pulses, such as indi cated at 44 and 45, curve j, to a pulse width discriminator 46.

The pulse width discriminator may be of any known construction capable of distinguishing between two or more different width characteristics. A. suitable circuit for this purpose is disclosed in the copending application of E.' Labin and applicant, Serial No. 487,072, iiled May 15, 1943, assigned to Federal Telephone and Radio Corporation. In Fig. 5, I illustrate such a circuit. The input pulses 44 and 45, which are of constant amplitude, are applied through a resistor 41 to an L-C circuit which is tuned to a frequency, the period of which is substantially twice the duration of the wanted pulse 45. The leading and trailing edges of the pulses applied to the tuned circuit shock excite the circuit so that when the desired pulse width isapplied, the shock effects of the leading and trailing edges thereof add accumulatively, as indicated at 48, 49 for pulse 44 and E0, El for pulse 45, curve k, and produce output undulationsas indicated at 42 and E3, respectively. For pulses of greater and lesser width than the wanted puise 45, the shock excitation effects of the leading and trailing edges thereof ido not occur exactly in synchronism with the half period of the tuned frequency so that the shock effect of the trailing edge opposes more or less the oscillatory energy established by the leading edge of the same pulse. Thus, the undulations resulting from pulses of greater o1- lesser width than the width of the wanted pulses are of less amplitude. than those obtained from the wanted pulses.

In order to prevent the trailing oscillations produced by one input pulse from interfering with oscillations established. by the next input pulse, a damping circuit E4 is provided whereby all oscilamplifier tube 81. .The tube 60 is so biased through resistor 10 as to pass a portion D1 of the larger amplitude pulses D occurring above a cutoil' level 7l, curve m, Fig. 7. A delay ti curve n, is imposed on the pulse portion Dr by means of the delay device 85 which preferably comprises a resistance-capacitance network. This network together with resistor 84 distorts the clipped energy as indicated at D2. The energy D2 being applied to the grid 66 alters the gain of the amplifier tube 61 which in this circuit is shown to be normally controlled by bias source 12, thereby distorting the input pulse D as indicated at Da, curve'o. The distortion, however, is in the trailing edge 13 of the pulse.

While the pulses Ds and E may be differentiated and clipped similarly as described in connection with the parts 28, 21, 36 and 31 of Fig. 1, I preferably subject the pulses to clipper gate 4i and pulseI width discriminator 48 which operate simi llarly as described in connection with Fig. 1. The

llations following the first undulation of positive polarity are eliminated. For amore thorough description of the pulse width discriminating circuit, reference may be had to the aforementioned application, Serial No. 487,072. y

The resulting undulations which correspond to the wanted pulses arethereafter separated from undulations produced from the unwanted pulses by a clipper E6 which clips and amplifies the peak portions extending above level 51 as indicated by output pulse 53a, curve L. It will thus be clear that a discrimination is made of the difference in width of the pulse portions obtained by clipping between limits 42 and 43. curve i, whereby undulations corresponding only to wanted pulses B are obtained.

Before applying the pulse output of mixer i5 to the clipper gate 4 I, it may be rst differentiated at 2B. I accordingly provide movable switch contacts 61 and/ 58, ganged as indicated at 59,

clipper limits the pulses between gate limits 42 and 43 to produce constant current pulses 14 and 15, curve p, the pulse 14 being narrower than pulse 15. By means of discriminator 46 pulse 18, curve q, is obtained from pulse 15, the unwanted -pulse 14 being eliminated. y

In Figs. 8 and 9 I show a further embodiment of the invention in which the delayed pulse portion is mixed with the unwanted pulse for dise tortion of the trailing edge thereof.k The unwanted pulse F of the pulse train 11, curve r, is applied to input connections 18 and 19, the energy over connection 'I8 being fed toan amplier 84 biased to threshold clip the pulses at a level-3i at least as high as the amplitude of the wanted pulse G, whereby the upper portion of the unwanted pulse F is obtained, amplified and inverted as indicated at Fi, curve s. The inverted pulse portion F1 is applied to a delay device 82 whereby the pulse portion is retarded an amount ta. It is then applied to a capacitance- resistance circuit 83, 84 where it is distorted as indicated at F2. The resistance 84 is chosen high so that it `operates to isolate connection 19 from the capacitance 83. The distorted pulse potential F2 is mixed with energy of the unwanted pulse F and the wanted pulses G at connection 85. Thepulse potential Fa being retarded by an amount t2 relative to the unwanted pulse F adds algebraically thereto as indicated by the wave q, Fig. 9. The leading edge B8 -of the pulse potential F2 subtracts from the trailing portion of the pulse F thereby distorting the trailing edge thereof as indicated at 81. The portion of the pulse potential F2 extending beyond the trailing edge of the pulse F appears in the wave at 88 as a negative pulse.

Here again, the wave train of curve t could be applied to the clipper and diiferentiator elements of Fig. 1 but instead I show the wave train to be applied to clipper gate 4| for clipping the pulses the pulses of which vary in width according to the v pulse portions clipped by gate 4I.

In Figs. 6 and '7, another system is shown including a triode 60 having an input v8| to which a'pulse train-82, curve m, may be applied. The

between limits ..42 and 43. This produces constant amplitude pulses 89, 90, curve u, which are applied -to width discriminator 46 whereby pulse `9i, corresponding to wanted pulse G .is obtained.

.While I have described the principles of my invention in connection with several specific systems. it will be clearly understood that this de-l scription'ismade'only by way of example. Many assesses other methods and systems oi distorting the larger amplittde pulses may be used. For exwhose leading or trailing edge, or both, are distorted. It will be understood, therefore, that the systems herein described are illustrative of the invention only and not restrictive of the scope ofthe appended claims.

I claim:

1. A method of discriminating between wanted and unwanted electrical pulses of substantially the same width and similar shape characteristics where the unwanted puise is of amplitude greater than the wanted pulse. comprising distorting said unwanted pulse to render it different in a shape characteristic other than amplitude from said wanted pulse, and thereafter processing said pulses to separate them according to their diierence in said shape characteristic.

2. A method according to claim 1 wherein said distorting operation renders at least a part of the unwanted pulse of a width different from the width of the wanted pulse, and said processing operation includes the step of discriminating between the width characteristics of said distorted pulse and said wanted pulse.

3. A method according to claim 1 wherein said distorting operation renders at least a part oi the unwanted pulse within the amplitude limits of said unwanted pulse of a width different from the width of the wanted pulse, and said processing operation includes the steps of clipping the pulses between two limits including said part of said unwanted pulse and a corresponding part of said wanted pulse, and eliminating the unwanted part by discriminating between the two parts according to their difference in width.

4. A method according to claim l wherein the operation of distorting said unwanted pulse includes decreasing at least in part the steepness of the slope of one of the edges thereof.

5. A method according to claim l wherein the operation of distorting said unwanted puise includes clipping energy of said unwanted pulse at a level at least as high as the amplitude of said wanted pulse, changing the shape of the clipped portion, and combining the clipped portion and the unwanted pulse.

6. A method according to claim l wherein the operation of distorting said unwanted pulse includes clipping the said unwanted pulse at a level at least as high as the amplitude of said wanted pulse, re-shaping the clipped portion to reduce the steepness vof at least one of the' edges thereof, and combining the re-shaped clipped portion and said unwanted pulse.

7. A method according to claim 1 wherein the f operation of distorting said unwanted pulse includes clipping said unwanted pulse at a, level at least as high as the amplitude oi said wanted pulse, retarding the clipped portion. and combining thevclipped portion and energy of said unwanted pulse.

8. A method according to claim 1 wherein the operation of distorting said .unwanted pulse includes clipping'said unwanted pulse at a level at least as high as the amplitude of said wanted pulse, retardingand reshaping the clipped portion to distort one of the edges thereof and to cause the distorted edge to overlap in time the trailing portion oi' the unwanted pulse, and com- 8 bining the clipped portion and energy o! said unwanted pulse.

9. A method o1' discriminating between wanted and unwanted electrical pulses ci substantially the same width and ot similar shape characteristics where the unwanted pulses are oi amplitude greater than the wanted pulses. comprising l producing, in response to said unwanted pulses.

electrical pulse potentials dissimilar to said unwanted pulses, combining the dissimilar pulse potentials and energy oi said unwanted and wanted pulses to produce a difference in a shape characteristic other than amplitude between the unwanted and wanted pulses, and thereafter processing said pulses to separate them according to their diierence in said shape characteristic.

10. A method according to claim 9 wherein the pulse potential producing operation includes producing a pulse potential difierent from the shape of said unwanted pulses.

11. A method according to claim 9 wherein the pulse potential producing operation includes producing the pulse potentials at a given time displacement With respect to the occurrence of said unwanted pulses.

12. A method according to claim 9 wherein the pulse potential producing operation includes producing in -synchronism with said unwanted pulses, pulse potentials having a shape dissimilar tothe shape of said unwanted pulses.

13. A method of selectively discriminating be' tween electrical pulses of different amplitude to eliminate those pulses of similar shape characteristics and of amplitude diierent from a given amplitude, comprising threshold clpping the pulses at a. level at least as high as the greatest amplitude of those pulses smaller than said given amplitude, distorting the pulses of amplitude greater than said given amplitude to render them different in a shape characteristic other than amplitude from saidwanted pulses. and thereafter processing said pulses to separate them according to their diiTerence in said shape characteristic. Y

14. A system for discriminating between wanted and unwanted electrical pulses of substantially the same width and of similar shape characteristics where the unwanted pulse is greater in amplitude, comprising means to distort at least one edge of said unwanted pulse to render it different in a shape characteristic other than amplitude from said wanted pulse. and means to process said wanted and said distorted unwanted pulses to separate them according to their difference in said shape characteristic.

15. A system according to claim 14 wherein the means to distort the unwanted pulse comprises means for decreasing at least in part the steepness of said one edge.

16. A system according to claim 14 wherein said shape characteristic is puise width and the processing means includes means to distinguish between said pulses according to their difference in width.

17. A system for discriminating between wanted and unwanted electrical pulses of substantially the same width and of similar shape characteristics where the unwanted pulse is greater in amplitude, comprising means to` clip the unwanted pulse at a level at least as high as the amplitude of said wanted pulses, means to change the clipped portion in a. shape characteristic other than amplitude portion, means for combining the clipped portion and energy of said unwanted pulse todistort the shape thereof in Magnan accordance with the .changed shape of said clipped portion, means to process the distorted unwanted pulse and said wanted pulse to separate them according' to their diilerence in shape.

18. A system according to claim 17 wherein the means for clipping said unwanted pulse includes a vacuum tube, and means to bias the vacuum tube to cut-ofi at a desired level.

19. A system according to claim 17 wherein the means for changing the shape oi the clipped portion includes a circuit having a condenser adapted to be charged and discharged by the clipped portion thereby altering the shape ot the leading and trailing edges thereof.

20. A system according to claim 17 wherein the means for changing the shape of the clipped portion comprises a circuit having a delay characteristic adapted to retard the clipped portion a predetermining amount so that said clipped portion overlaps in time the trailing portion oi.' said unwanted pulse.

2l. A system according to claim 17 whereinthe means for combining said clipped portionand energy of said unwanted pulse includes a mixer tube, and means for varying the gain of said tube in accordance with the potential ofsaid clipped portion.

22. A system according to claim 17 wherein the means for combining said clipped portion and energy of said unwanted pulse includes means for algebraically adding said clipped portion to the energy of said unwanted puise.

23. A system according to 'claim 17 wherein said shape characteristic is pulse width and said means for processing said distorted unwanted pulse and said wanted pulse includes means for clipping the pulses between two limits including a distorted part of said unwanted puise and a corresponding part oi' said wanted pulse, and width discriminating means for eliminating the unwanted part.

DONALD D. GRIEG.

REFERENCES CITED The following references are of record in the ille of this patent:

UNITED STATES PATENTS Number Name Date 2,153,202 Nichols Apr. 4, 1939 2,157,677 Runge May 9. 1939 2,180,355 Haicke Nov. 21. 1939 2,212,173 Wheeler et al. Aug. 20, 1940 2,310,692 Hansell Feb; 9, 1943 2,326,584 Van Zeist Aug. 10, 1943 2,241,170 Uibricht May 6, 1941 2,286,450 White et al. June 16, 1942 2,294,341 Moore Aug. 25, 1942 2,344,697 Hollingsworth Mar. 21, 1944 FOREIGN PATENTS Number Country Date 290,979 Great Britain May 18, 1928

Images