US2434937A - Selective filtering system - Google Patents

Description

'double triode tube.

-as a normal amplier.

34 effectively Yby-passes the mixer stage 39.

3 D. D. Grieg Serial No. 459,959, filed September 28, 1942, now Patent No. 2,416,306, granted Feb. 25, 1947. The signal increments thus converted from time modulated pulses to amplitude modu lated pulses are readily applicable in the usual manner to an audio stage for audible reception. Referringtoliigs. 2 and 3 -cf the drawings, curve a of Fig." 3 represents an input wave or train of pulses of various shape characteristics. Pulse I is a wanted signal pulse which appears at a given repetition rate and is time displaced by intelligence being transmitted. Pulses II and I2 dii'rer in amplitude from the pulse I0, the pulse II being less and the pulse I2 being greater. These two pulses are selected of the same width as pulse I0 for purposes of illustrating the selective feature of the amplitude discriminator 1. Pulse I3 is of the same amplitude as pulse I0 but is of greater width. Pulse I4 is also of the same amplitude as pulse Ill-but is of different shape in that the leading and trailing edges are sloped at different' buildup'and decay times.

This train of pulses after having passed through the detection unit 4 is applied to the interference limiter 5. The input I5 of the amplitude discriminator 1 feeds the pulses to a threshold clipperV I6 which preferably is adjusted 'to clip the pulses at an amplitude level I1, curve a, Fig."3, at an'vamplitude level slightly greater than the amplitude of the small pulse II but less `than= the amplitude of pulse I0. This clipping `the larger pulse I2. The output of the clipper stage 22 is taken from the plate 'connection 24 so that the upper portion of the pulse thus clipped is inverted in phase and amplied as indicated at I2aa. This inverted energy I2a is applied-to an input connection 2E of grid 3| of a mixer stage 30 through a movable switch contact 25 and a terminal 21.

The mixer stage 35 comprises a conventional While the plate output of the clipper 22 is applied to grid 3| of the mixer, lthe outputenergy of clipper IB is also applied to the mixer through a connection 32 and switch .34 to' grid 33.- The output at the anode connecjtion 36 is illustrated by curve c.

served that the large pulse I 2 as well as the small It Will be 0bpulse I I is eliminated. This elimination may result in some smallmdisturbance I 2b but this. as will be seen hereinafter, will be eliminated by the operation of a subsequent filtering stage.

low amplitude' level I1 curve a, there is no need Aof the clipping operation of tube 22 or the mixing of mixer stage 30.' The switch arm 25 may then be moved to its upper contact, opening switch 34 and shifting the bias of tube 22 to operate The opening oi.' switch Curve c, however, still contains pulse I3 which jdiiers in width from pulse I 0 and pulse I4 which-differs in slope from pulse I0. It will be noted that the base of the pulse `I4 is substantially the same width as pulse I 0. The pulse width having slower build-up and decay times such as the pulse I4. Should it be desirable, however, to eliminate pulse I0 in preference to pulse I4 a re-shaping of pulse I4 will then be desirable. This re-shapng is accomplished by the gate function of the slope discriminator 8, the details and operation. of which- 5willebefhereinafter described. Fonthe presentlet it be assumed that the device 8 is so adjusted as to pass the pulses I, I3 and I4 substantially as shown in curve c.

The width selector 9 is preferably of the character disclosed in our copending application Serial No.` 487,072 filed May l5, 1943, entitled Pulse selecting and eliminating system. Other forms' of width selector that may be used are disclosed lin the copending application of E. Labin Serial No. 467,509, led December 1, 1942, now Patent No. 2,418,127, granted April 1, 1947.

The output connection 35 of the shape discriminator 8 is applied to a limit clipping tube 42 which limits the amplitude of the pulses I0, I3 and I4 to substantially the same amplitude. This tube also operates to invert the pulses from positive to negative as indicated by curve d. This output energy o f tube 42 is applied through a resistor R to a shock excitable L-C circuit 45. The condenser C is adjustable so that the circuit may be tunedtoa frequency the period of which is twice or some other proper multiple of the duration of the Wanted pulse I0. The coil L, of course, may also be made adjustable either in place of the adjustment of condenser C or to operate V'in conjunction therewith.

Connected across the tunable circuit 45 is a vacuum tube 46 the cathode 41 of which is connected to the input side 43 of the circuit 45 while the anode 48 is connected to the opposite side 44 of the circuit. The anode connection at this side of the circuit is also connected through lead 49 to a positive potential B+. A condenser 50 separates the lead connection 49 from a ground lead 52. The condenser 50 operates as a coupling condenser and the capacitance thereof does not affect appreciably the adjustment of-the condenser C.

The energy output of the anode connection 4I of the tube 42 is also applied to a grid 54 of the tube 46 so as to block conduction between the cathode 41 and the anode 48 while pulse energy is applied to the circuit 45. The undulations occurring in the circuit 45 are taken oir through a connection 53 for application to a threshold clipper and amplifier stage 60. The bias on the grid 6| of the clipper stage 6D is controlled by adustment of a resistor 62.

The operation of the width selector portion of the circuit of Fig. 2 is indicated by the curves d, e and f. The curve d represents the anode voltage across the resistor R. Curve e represents the oscillations produced at the output 56 of the cir- Acuit 45 in response tothe pulse energy of curve d. Curve f represents the pulse output at 64 from the clipper stage 60 after the undulations have been thresholdclipped.

When the leading edge 1I of the wanted pulse I0 is applied to the circuit 45, an undulation 12, curve e is produced. The circuit 45 will oscillate tending to produce in response to the edge 1I a damped wave according to the undulations 12, 13, 14, etc. When the trailing edge 15.01 the pulse I0 is applied to the circuit 45, a further shock excitation 'occurs in the opposite direction producing undulations 16, 11, etc. The undulations 13 and 16 are in time relation since the circuit is tuned to aperiodcqual to twice the dura- 'engager 7 probability of identical amplitudes'is very small, 'and furthermore it will be understoodthat the pulse shape used for conveying messages may be varied from time to time so that the enemy will be unable to effectively jam the system.

It will also be recognized that the i'lltering system of our invention may be used in conjunction with multi-channel telegraphy as well as in a radio receiver.

While we have shown and described the prin ciples of our invention in connection with specic apparatus, we recognize that various changes and modifications may be made therein without departing from the invention. It is our aim, therefore, to cover in the appended claims all such changes and modifications as fall within the scope of the invention.

We claim:

1. A method of selectively illtering a train of pulses to eliminate those pulses differing in ampitude and width characteristics from a given pulse shape, comprising threshold clipping the pulses at an amplitude greater than the smaller amplitude pulses but less than` the amplitude of said given pulse shape, thereby eliminating. said smaller pulses, threshold clippingand inverting in phase at least a part of that portion of those pulses extending above the amplitude of said given pulse shape, mixing the inverted pulse energy with the pulse energy obtained bythe firstmentioned clipping operation,` whereby .those pulses of amplitude greater than said given pulse shape are eliminated, shock exciting aresoriant circuit by means of the leading and trailing edges of the remaining pulses, the oscillations initiated by the edges of each pulse combining to form an undulation following the trailing ,edge o f the pulse, damping out the oscillations normally Vfollowing said undulation so that oscillations produced by one pulse will not affect the oscillations produced by the next pulse, tuning the resonant circuit to a period equal to a given multiple ofthe width of the given pulse shape so that vthe oscillations set up by the edges of the pulses of width corresponding to the width of'said given pulse shape provide undulations of maximum amplitude and the oscillations produced by the 'edges' of pulses of greater and lesser width form undulations of amplitude less than said maximum undulation, and threshold clipping the undula' tions to eliminate those corresponding to those greater and lesser width pulses.

2. A method. of selectively filtering 'a train of pulses to eliminate those pulses diiering in widthl and slope characteristics from a given pulse shape comprising gate clipping to obtain'a slice lof each pulse at a level where the given pulse-shape differs in width from the widths of the other pulses, shock exciting a resonant circuit with the pulse slices, the oscillations initiated by the edgesofv each pulse slice combining to forinzanundulatiori following the trailing edge 'of "the, pulse slice, damping out the loscillations normally following d spending to those greater and lesser widthpulse slices.

3. A system 'for selectively filtering a train of pulses to eliminate those pulses differing vin amplitude and width characteristics from a given pulse shape, comprising a rst means to threshold clip the pulses at an amplitude greater than the smaller amplitude pulses but less than the amplitude of said given pulse shape, a second means to thereshold clip and invert in phase at least a part of that portion of those pulses extending above the amplitude of said given pulse shape, a mixer, means to supply the output of said first and second clipping means to said mixer so that the inverted pulse energy substantially eliminates the larger pulses, a pulse width discriminator, and means to apply the output of said mixer to said width discriminator thereby eliminating those pulses diiering in Width from said given pulse shape.

4. The system defined in claim 3 wherein the width discriminator comprises a shock-excitable resonant circuit responsive to the leading and trailing'edges -of the pulses to establishloscillations, means to damp out the oscillations normally following the first undulation formed by the combined oscillations, tuning the resonant circuit to a period equal to an even multiple of the width of the given pulse shape so that the oscillations set up by the edges of the pulses of Width corresponding to the width of said given pulse shape provide undulations of maximum amplitude and the oscillations produced by the edges of pulses of greater and lesser width form undulations of amplitude less than said maximum undulation, and means to threshold clip the undulations to eliminate those corresponding to the greater and lesser width pulses. Y

`5. The system defined in claim 3 wherein the means for conducting the pulses from the mixer v to the width discriminator comprises a gate clipsaid undulation so that oscillations produced by e one pulse will not affect the-oscillations produced by the next pulse slice, tuning the resonant cir-5 cuit to a period equal to a given multiple of the width `oi the slicetaken from the pulses of; said given pulse shape so that the oscillationsvset un by the edges of pulse slicesof greater and lesserimum amplitude and the-oscillations produced by the edegs of pulse slices of greater and lesser 'Width form undulations of amplitude less thansaid maximum undulation, and threshold clipping the 'undulations toeliminate t1- io se corre ping device, means to control the breadth of the gate, andmeans to control the level at which the gate clipping operation is performed on the pulses whereby said given pulse shape may have given slope characteristics and the pulses diiering therefrom are eliminated. 6. A system for selectively filtering a traino! pulses to eliminate those pulses differing from one or more of the given amplitude, given slope and given width characteristics of a given pulse shape, comprising filter means to pass energy of those pulses having said given amplitude characteristic, filter means to pass energy of those pulses having said given slope characteristic and to translate them into pulses of said given width, filter means to pass energy of those pulses having said given width characteristic, and means to correlate the operations -of the several iilter means to produce nal output pulses having a timing according to those ,pulses of said train having said given shape characteristics, said means for correlatingV the operations of the several iilter means including means for connecting said filter means in tandem relation whereby the width iilter4 means operates on the pulse energy output of the other filter means.

7. A system for selectively iiltering a train of pulses to eliminate those pulses differing from one' or more of the given amplitude, given slope aridi given width characteristics of a given pulse' shape, comprising first iilter means to pass energy' of those pulses having said given amplitude characteristic and to translate the slope characteristic into a width characteristic, and second lter means including a resonant circuit connected to the output of said rst lter means to pass pulse energy of those pulses passed by said first filter means having said given width characteristic.

8. A system for selectively filtering a train of pulses of substantially the same base width to eliminate those pulses differing in slopes of the edges thereof from the slopes of the edges of a given pulse shape, comprising means for clipping the train of pulses between two selected clipping levels relative to the base of the pulses to obtain a slice of each pulse, the pulses differing in slope from said given pulse shape producing pulse slices differing in width from the width of the pulse slices obtained from pulses of said given shape, a pulse width discriminator, and means to apply the pulse slices from said clipping means thereby eliminating those pulse slices differing in width from the pulse slices clipped from the pulses of said given pulse shape.

9. A method of selectively ltering a train of pulses to eliminate those pulses diierlng from one or more of the given amplitude, given slope and given Width characteristics of a given pulse shape, comprising subjecting energy of the train of pulses to a first filtering operation to pass energy of those pulses having said given amplitude characteristic. subjecting the pulse energy output of said rst filtering operation to a second filtering operation to pass pulse energy between two selected clipping levels to obtain pulses of duration different from said given width characteristio for those pulses varying in width and slope from said given shape, and subjecting the pulse energy output of said second filtering operation to a third filtering operation to pass energy of those pulses having said given width characteristic.

10. A method of selectively filtering a train of pulses of substantially the same base width to eliminate those pulses differing in the slopes of the edges thereof from the slopes of a given pulse shape, comprising: gate clipping at a given level above the base of the pulses to obtain a slice of each pulse, the pulses differing in slope from said given pulse slopes producing pulse slices differing in width from the widths of the pulse slices obtained from pulses of said given shape, and subjecting the pulse slices to a ltering operation to eliminate those pulse slices differing in width from the puse slices clipped from pulses of said given pulse shape.

11. A method of selectively filtering a train of pulses to eliminate those pulses differing in one or more of the folowing characteristics; amplitude, slope and width characteristics from a given pulse shape comprising: threshold clipping the pulses at an amplitude greater than the smaller amplitude pulses but less than the amplitude of said given pulse shape, thereby eliminating said smaller pulses, threshold clipping and inverting in phase at least a part of that portion of those pulses extending above the amplitude of said given pulse shape, mixing the inverted pulse energy with the pulse energy obtained by the rstmentioned clipping operation, whereby those pulses of amplitude greater than said given pulse shape are eliminated; then eliminating pulses having diiTerent slopes than Vthat of said given pulse including gate clipping to obtain a slice of lesser width pulses.

each of the remaining pulses at a given level with respect to the bases of the pulses, the pulses differing in slope from said given pulse slopes producing pulse slices differing in width; and thereafter filtering the pulse slices to eliminate those pulse slices differing in width from the pulse slices taken from pulses of said given pulse shape.

12. A system for selectively filtering an original train of differently shaped pulses to produce a new train of desired pulses corresponding to a given shape comprising: means to eliminate those pulses of undesired amplitude, means to eliminate those pulses of undesired slope by changing their Tease width from that base width of the desired pulses, means coupled to the output of the lastnamed means to eliminate those pulses of undesired width, and means to correlate the results from the above-mentioned means to produce said new train of desired pulses,

13. A method of selectively ltering a train of pulses to eliminate those pulses differing from one or more of the given amplitude, given slope and given width characteristics of a given pulse shape, comprising: subjecting energy of the train of pulses to a first filtering operation to pass energy of those pulses having said given amplitude characteristic, subjecting the pulse energy output of said rst filtering operation to a second ltering operation including a clipping operation to produce pulse energy between two selected clipping levels to obtain pulses of different duration for those pulses varying in width and slope from said given shape, subjecting the pulse energy output of said second filtering operation to a third filtering operation to pass energy of those pulses having said given width characteristic, said third ltering operation including shock exciting a resonant circuit by means of the leading and trailing edges of the pulses from said second ltering operation, the oscillations initiated by the edges of each pulse combining to form an undulation following the trailing edge of the pulse, clamping out the oscillations normally following said undulation so that oscillations produced by one pulse will not aiect the oscillations produced by the next pulse, tuning the resonant circuit to a period equal to a given multiple of the width of the given pulse shape so that the oscillations set up by the edges of the pulses of width correspending to the width of said given pulse shape provide undulations of maximum amplitude, and the oscillations produced by the edges of pulses of greater and lesser width form undulations of amplitude less than said maximum undulation, and threshold clipping the undulations to eliminate those corresponding to those greater and EMILE LABIN. DONALD D. GRIEG.

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

UNITED STATES PATENTS Number Name Date 2,151,149 Poch Mar. 21, 1939 2,211,942 White Aug. 20, 1940 1,483,172 Gannett Feb. 12, 1924 2,284,714 Bedford June 2, 1942 2,132,655 Smith Oct. 11, 1938

Images