US2562660A

US2562660A - Pulse generating circuit

Info

Publication number: US2562660A
Application number: US512931A
Authority: US
Inventors: Chance Britton
Original assignee: Individual
Current assignee: Individual
Priority date: 1943-12-04
Filing date: 1943-12-04
Publication date: 1951-07-31
Anticipated expiration: 1968-07-31

Description

B. CHANCE PULSE GENERATING CIRCUIT Filed Deo. 4, 1943 July 31, 1951 vthe invention proceeds.

Patented July 31, 1951 PULSE GENERATING CIRCUIT Britton Chance, Cambridge, Mass., assigner, by mesne assignments, to the` United States of America as represented by the Secretary of the Navy Application December 4, 1943, Serial No. 512,931

(Cl. Z50- 27) 4 Claims.

This invention relates to pulse-forming devices and particularly to such devices utilizing thermionic tubes.

One of the objects of the invention is to provide a pulse-forming device in which the time duration or Width of the pulse produced may be accurately controlled.

Another object of the invention is to provide a pulse-forming device in which the time duration or width of the pulse produced varies linearly with an applied voltage.

Another object of the invention is to provide a pulse-forming device of the delayed multivibrator type in which the time duration or width of the pulse produced varies linearly with the ner of connecting the various components of the circuits will be apparentas the description of The invention has been illustrated in the accompanying drawings in which,

Fig. 1 is a circuit diagram of a pulse-forming device embodying the invention;

Fig. 2 is a diagram showing the grid voltage of one of the tubes plotted against time during a certain portion of the cycle of operation of the circuit;

Fig. 3 is a circuit diagram of a modified for of the circuit which provides compensation for temperature changes; and

Fig. 4 is a circuit diagram of another modified form of the invention with which evengreater stability may be obtained. l A 'f In carrying out the objects of the invention I may use two thermionic tubes, each having at least a plate, a grid,v and a cathode, andthese maybe incorporated in a single envelope I, as shown in Fig. 1. The tube I I within the envelope III comprises a plate I2, a control grid'll'l, anda cathode Ill, while the tube I5. comprises; a plate I6, a control grid Il, and a cathode I8. .The control grid I3 of tube II may be given any desired bias potential through a resistorl and a' potentiometer 2 I, across which a source Aof voltage, as, for instance, a battery 22,` maybe connected.

The cathodes I4 and I8 of tubes I'I and I5 are shown connected together and to groundv through -a common cathode resistor 24, while the heaters for these cathodes together with the energizing circuits have been omitted in the drawing for ciearness.

The plates I2 and I6 are connected respectively through

resistors

25 and 26 to a source of positive potential, indicated at 2l. The grid I'I of, the tube I5 is also connected through a resistor 28 to the source of potential 2'I. This resistor 28 is preferably a great deal larger than the

resistors

25 and 26 and is made large to prevent excessive grid current in the tube I5. The plateY I2 of the tube II is connected to the grid I'I of the tube I5 through a condenser 29. The pulse produced by this device may be obtained by connecting a suitable output circuit to the plate I6 of the tube I5 or to the plate I2 of the tube II, as indicated respectively by the wires SI and 32. Y

An initiating or trigger pulse may be applied to the grid I3 of the tube II through a condenser 3l) from a sourcenot shown. This trigger pulse must be a sharply peaked positive voltage pulse of very short time duration to insure accurate width or time duration of the pulse to be produced. The values of the condenser Si) and the resistor 20 are so chosen that these components function as a differentiating circuit to aid in producing this type of trigger pulse. Some means, as, for instance, the diode I9, should be provided- Yto prevent negative voltage pulses from being applied to the grid I3 of tube I I during the time the circuit is operating to produce the desired pulse. Alternatively a negative trigger pulse may be applied to the grid I1 of the tube I5.

In the quiescent condition of the circuit of the invention the tube I5 is conducting, since the grid I1 thereof is held at approximately the potential of the lcathode I8 by the grid current through the large resistor 28. The grid I3 of tube IIl on the other hand, has sufficient-negative bias potential applied to it normally to prevent this tube II from conducting. In this condition the cathodes I li and I8 are maintained above ground potential by the flow of plate current of the tube I5 through the commoncathode resistor 24. This maintains the grid I3 of the tube I I at a sufciently lower potential than the cathode I4 to keep this tube shut oil. At this the grid I'I of tube I5, which is at a lower potential determined by the grid current in the resistor 28.

If nowV a positive trigger pulse of sufficient amplitude is applied to the grid i3 of the tube I I through the condenser 3G this tube will start to conduct. When this conduction begins, the potential of the plate I2 of the tube II Will be lowered by an amount equal to the voltage drop across the resistor 25. This drop in potential is applied to the grid I'I of the tube I5 through the condenser 2S, and must be suflicient in amplitude to cause the tube I to cease conducting. When the tube S5 ceases to conduct the potential of the cathodes I4 and I8 is determined by the plate current of the tube II through the cathode resistor 24. The bias potential of the grid I3 mustnow be sumciently positive with respect to the cathode I4 to maintain the tube i I conducting.

A trigger pulse with a steep front is desired to cause theI tube i5 to cease conducting and the tube i i to begin conducting with a minimum time delay after the trigger is applied to the vgrid I3 of the tube II. With the proper trigger, this action Will takeiplace in a fraction of a-microsecond.

During the quiescent Astate the condenser 29 had been charged by the diference in potential between th-eplate A*I2 4of the tube I I -an'd the grid i1 of the tube I5. When the tube I I starts to conduct, the voltage drop in the resistor 25, caused by the plate current of the tube `I I, is also applied to this condenser and is opposite in potential to the charge thereon. The change in voltage of the plate I2, acting through the condenser 29, drives the grid lII of ythe tube I5 to va potential withrespect'to thecathode I8 which'is suncient tocut-oi the tube I5, and the discharge path of this condenser 29 is now through the'resistors 28 and '2.5. But the plate current of the tube I I, while this tube is conducting, causes a constant potential drop across the resistor 25. Therefore, during the time that the tube I5 is cut-oil, the condenser 2E) is effectively in series with a source of constant potential and the resist-or 28. Since the grid I'I of the tube l5 is connected by the resistor 28 to the source of voltage at 21, the potential of the grid I! tends to rise to the voltage at 21 to establish equilibrium conditions'in this vseries ycircuit consistingof the condenser 29, the resisground. This-causes a reduction of the vcurrent in tube 1I and an increase in the voltage of the plate I2. Through the condenser 29, this plate voltage rise causes the potential of the grid I1 to rise further in potential, increasing thecurrent in tube I5 which further raises'the potential of Vthe cathodes I4 and I8. Thus. the action is cu- 1nulative,'the tube I5 rapidly .building up to its quiescent-conducting state and the tube II being cut off. The grid current in the tube I5 then holds the grid Il' at substantially cathode potential, and the condenser -29 starts to charge again.

The tendency ofthe grid vI'I to assume the potential of the source Yat V2l during the condenser discharge is what makes this circuit capable of producing pulses of `ver-y 'accurate width. The

curve 33 of Fig. 2 shows the variation with time of the voltage 'of the grid I'I during the period that tube I5 is cut-oi.

Grid voltage is plotted vertically from the point 34, which represents the most negative potential 4 of the grid l?, as produced by the operation of the tube II, to the point 35 which represents the potential of the source 21. As indicated, the cutoif grid voltage of the tube I5 is represented by the dotted line 36. As the potential of the grid I1 rises exponentially with time with the discharge of the condenser A29, the'voltage necessary Yfor the tube I5 "to begin conducting (which is the cut-01T voltage) will be approached during the time that the rate of increase of the voltage of the grid Il is substantially its maximum. This insures the tube I5 coming on and the tube I I being cut-off at a Avery accurate time interval after the initial trigger pulse is applied to the grid i3 of the tube I i.

In the usual arrangement of multivibrator circuits the grids derive their biasing potential from the cathodes through suitable grid resistors. In such a multivibrator when one grid is driven more negative than the cathode through the coupling condenser which is connected to the plate circuit ofthe Vother tube, the potential of the grid will tend to rise, as the condenser discharges until it reaches the potential of the cathode. 'If such a grid were drivensuiiiciently negative to reach the point 34 in Eig. 2, and then Were permitted torise in potential as the coupling condenser discharged, the rise in potential might follow a curve some- Whatsimilar to the curve 38, shown in Fig. 2. Here, as the condenser discharges the potential of the grid will approachzero or cathode potential. It Will Ybeseen that as the voltage of the grid vapproaches the cut-off potential of the tube the angle it makes with the cut-off potential line 36 at the point of crossingwill be so small that any slight change inout-off potential will cause a relatively large variation in the length of time that the particular .tube is non-conducting. With the vcircuit .of the invention, the .much steeper intersectionof theg-rid-voltage curve 33 with the line 36 `insures a very small time .difference for variations in the cut-off :voltage and therefore produces 'a'much more accurately timed pulse.

The "Width .or time duration ofthe pulse produced bythe circuit ofthe invention may be controlled in several ways. It has been pointed out that the `Width of the pulse produced in the plate circuit of the `tube I5 is ,determined by the :time .required ifor the condenser 29 to discharge to the point Where'the grid I1 is positive ienough with respect'to the cathode I8 to start 'the tube I5 conducting. This time is dependent .not nnly.onthe charge ofthe condenser 29 but also on the potential of the cathode I8. In the .quiescent .state the condenser 29 is charged by the ldifference Vinnpotential between the plate I2 and the grid I1. The conduction ofthe tube II 'also impresses a. voltage drop on the condenser'29. The magnitude 4of this voltage drop in the resistor 25 may be controlled by the .bias on 'the grid I3 of the tube II, since this bias voltage determines the plate current when the tube II is conducting. This bias voltage also determines the potential of the cathodes I4 and I8 vduring the time the tube II is conducting. and therefore governs the vamount condenser 29 must discharge before the tube I5 begins conducting.

vIt will be seen from 'these statements that the time 'the tube I5 is non-conducting which 'time vis-equal tothe pulse-width, maybe changed by varying either the value of the condenser '29, the Avalue of the resistor 28, the value of the resistor VV24,-'or the'bias potential applied to the grid I3. One vor more of these values may es be changed to make the desired adjustment. but probably the most convenient method is to adjust the bias on the grid I3 of the tube II.

An important feature'of this invention is that with the correct choice of circuit constants the 24 and 25 are equal in value, the percent devia-2l tion from linearity at the maximum pulse Width would be about 1% of one percent. It has been Y found experimentally, however, thatsubstantially as good results are obtained when the value of the cathode resistor 24 yis increased 'or decreased Vby as much as 500 ohms. Therefore these values are notcritical.

In one instance the circuit of Fig. 1 was' used with a double triode tube of the GSNTI type. The

voltage of the source 21 was A250 volts; the re` sistor was 25000 ohms; the resistor 24 vwas 3000 ohms; the resistor 25 was 5000 ohms; the resistor 26 was 5000 ohms; and the resistor 23 was 3 megohms. The values of the

condensers

30 and 29 were 5 and 300 micro-micro-farads,V

respectively. These values gave aA cathode potential of 42 volts in the quiescent condition. When the tube Il was conducting and the tube I5 nonconducting, the cathode potential dropped to 25`volts. lBy changing the bias .voltageonthe grid I3 of the tube II in this particular l,exarrlple from 19.4 volts to 35.3 volts the pulse .width was changed linearly with the bias voltage from 1.832 microseconds to 110 microseconds.

By using resistors and condensers in the eircuit of Fig. 1 that are little "eiected byv temperature variations, accurate pulse widthsr will be obtained over long periods of operationi-How obtaining temperature compensation it such is desired. The circuit is similar to that shown in circuitof the grid I1. The cathode 41 is connectedto the source of voltage 21 by a resistor 48 and to ground through a resistor 49. The plate 46 of the tube 45 is connected directly to the grid I1 of the tube I5, and through a resistor 50, which replaces the resistor 28 of Fig. 1,to the source of potential at 21. The potential of the plate 46 'of the tube 45 and therefore the potential ofthe grid I1 of the tube I5 may be l0` xed by `adjusting the

resistors

50, 48, and 49. These resistors must be adjusted so that during the time that the tube I5 is conducting, the

vvgrid I1 of this tube is slightly negative with respect to the cathode I8. IIf the plate 46 of thetube '45 is maintained very slightly positive with respect tothe cathode 41 a negative voltage pulse applied tothe grid' I1 of the tube I5 through the condenser 29 will cause the tube 45 to cease conducting. The discharge path of the condenser 29 willthen be through the resistors and 25, as previously described, in

connection with Fig. 1. The resistor 50 in the circuit of Fig. 4 serves the dual purpose of determiningA the time required for the condenser 29 to discharge sufficiently for the tube I5 'to begin conducting'and also, by means of the resistors v48 and 49, 'ofmaintaining'the grid I1 f at the desired potential to prevent grid current lfrom flowing.

The modication of the circuit of Fig. 1 in this manner eliminates any variation in pulse width that may result because of the tube I5 drawing which, of course, mayv be omitted, if desired,

from the circuit of Fig. 4, the accuracy of the pulse width obtained is greatly increased over "tubes by means of which a pulse of a given width or time duration may be produced very accurately by adjusting certain components of the circuit, such as a grid bias voltage. In addition, the width or time duration of this pulse may be made to vvaryflinearl'y with the grid bias poten- Fig. 1 except that the common cathoderresiston- 24 of Fig.l 1 has been replaced by two resistorsy 40 and 4I one of which, as, for instance, the resistor 40, is a resistance having a positive temperature coefficient of resistance, such as a wire wound resistor, and the other, as, for instance the resistor 4I, is a resistance having a negative temperature coeiicient of resistance, such as a carbon resistance. As far as the pulse-forming operation of the circuit is concerned, the purpose of these two resistors is the same as that of the resistance 24 in Fig. 1.

However, when a proper choice of values for the resistors 40 and 4I is made, so that the proper variation of resistance with temperature changes is obtained, eiects of temperature on the circuit may be largely eliminated.

Under certain conditions variation of the pulse width may result because the grid I1 of the tube I5 of the circuit of Fig. 1 draws gridcurrent during the time that the tube I5 is conducting. In order to prevent the flow of grid current, the circuit of Fig. 1 or Fig. 3 may be modified, as shown in Fig. 4. In this figure the tubes II and I5 are connected as in Fig. 3, but in addition a thermionic tube 45, having at least a plate 46 and a cathode 41 is connected in the tial. If properly constructed, the linearity of the pulse width variation may have an accuracy of one tenth of one percent (0.1%).

Various modications may be made in the in vention without departing from the spirit thereof, and I do not therefore wish to limit myself to what has been shown and described except as such limitations occur in the appended claims.

What I desire to claim and secure by Letters Patent is:

1. A pulse-forming circuit comprising, in combination, a rst thermionic tube having at least a plate, a grid, and a cathode, a second thermionic tube having at least a plate, a grid, and a cathode, a common cathode circuit for said tubes. a pair of resistors in said cathode circuit, one having a positive temperature coeicient of resistance and the other having a negative temperature coenicient of resistance, the values of said resistors being so proportioned that temperature changes in said circuit are compensated for, a source of positive potential, a third resistor connected between said source of positive l potential and the plate of said first tube, a fourth resistor connected between said source of positive potential and the plate of said second tube, a fifth resistor connected between the grid of said second tube and said source of positive ra-,temeon -of common cathoderresistors,the values ofsaid nfth, fsixth, and seventh resistors .-beingsochosen `that .thegrid of vsaid secondathermionictubeis normally maintained at -such a potential .with `respect to the cathode sthereof as u,to vmaintain lsaid second tube conductive but to -prevent the ow Vof grid 4current'therein when :said tube is conducting and .to ,cause said -grid to tend to reach a potential much higher than `the cathode 'of said second tube 4,when-said tube stops conducting, m'eans to apply va biasing potential to the vgrid -of said -first tube, so `:as normally -to maintain lsaid .tube :non-conductive when Ysaid second tube is conductive, land means to ,apply a positive initiating pulse to the grid o'f said first tubetocause said tube to becomefconductive.

2. -A -pulsefforming :network comprising, first =and vsecond electron tubeseach having at least an anode, a cathode, and a controlgrid,:a source of anode potential, :means including a resistance connecting each of said Vanodes to said source of anode potential, a common cathode circuitffor Vsaid tubes, a condenser coupling the -anode of saidiirst tube to the control Igrid of said second tube, means 'including an adjustable Ysource of potential coupled .to the grid of said rst tube for controlling the bias thereof, means -resistively coupling the grid of said second tube to Ysaid source of anode potential `for maintaining .said :gridat a positive potential, and ymeans :including a diode coupling'thesgrid ofsaid second tube to a positive potential less Athan the potential of .said source, the anode of said diode being connected to the grid ofsaid second tube.

3. A pulse-forming network comprising, first and second electron tubes Yeach havingrat least lan anode, a cathode and a control grid, 'a source of anode potential, lmeansincluding a resistance connecting each of :said anodes to said source of anode potential, a common cathode circuit vfor said tubes, a -condenser coupling the anode of said irst .tube to .the control y'grid of said-second tube, .means including an adjustable source .of

potential-coupled tothe `grid of Vsaid first tube .for .controlling the bias thereof, :first .coupling meansresistively connecting the'grid of said sec ond tube to said :source of :anodepotentiaL and secondcoupling means includinga diode-and a resistor `:serially connected rbetween fthe grid of said second tube -andsaid source of anode fpotential, .the anode ,of said diode being connected to ,the grid'of said second tube.

4. A 4.pulse-.forming network comprising. in

- combination, first and. second electron t-ubesv each having atrleast fan anode, a cathode, anda control:grid, asource of anode potential, means .including a iresistance `connecting each .of said anodesto said ysource of.anode potential, a common-cathode circuitincludinga.resistor for said tubes, `means including aresistance connecting the .control grid of .said second tube to said source of anode potential, a rst condenser v.coupling the anode oflsaid rst-tube to the control -gridof said second tube, fa-source-of A.biasing po- .tential, va potentiometer vconnected across said source .of .biasing potential, means including a vresistance and .a serially connected diode cou- ,pling 4said ,potentiometer -to .the control .grid .of

said first tube, the cathode of said .diode .being connected to -the gridof said rsttubaasecond .condenserconnected-to the anode .of said diode, .andrmeans for .applying a positive .pulse to Vsaid second condenser.

BRIITON CHANCE.

REFERENCES 'lCITED The following references are o'f record 1n the Iile of this patent:

EUNrrED STATES PATENTS Number .Name :Date Re.;19',735 Lederer Oct. 22, '1935 :2,102,951 lIiIackenberg Deal-21, 1937 Y2,185,363 `White Jan. 2, 1940 2,185,367 Blumlein Jan.2, 1940 12,207,511 Geiger July 9, y1940 12,234,690 Depp Mar. 11, 1941 2,265,996 Blumlein Dec. :16, 1941 2,356,761 .Jionesetal Aug. 29, 1944 12,405,237 Ruhlig Aug. 6, 1946 2,428,058 Wise Sept. 30, 1947