US2949578A - Millimicrosecond pulse circuits - Google Patents

Description

Aug. 16,1960 J.A.NARUD 2,949,578

MILLIMICROSECOND PULSE CIRCUITS Filed Jan. 29, 1957 2 Sheets-Sheet 1 IN V EN TOR.

Aug. 16, 1960 J. A. NARUD 2,949,578

MILLIMICROSECOND PULSE CIRCUITS I Clan Arne /Varud BY y( i du 0 WW 4 r rom/E Ys lnited States llatent Zdg Patented Aug. 16, i960 ice MILLIMICROSECOND PULSE CIRCUITS Jan A. Narnd, Concord, Mass., assigner to the United States of America as represented by the Secretary of the Navy 'Filed llan. 29, 1957, Ser. No. 637,051

4 Claims. (Cl. sns- 34) The present invention relates generally to electronic pulse circuits and, more particularly, to apparatus for and methods of generating, discriminating and counting millimicrosecond pulses.

In the fields of nuclear instrumentation, computers, automatic controls, etc., the need has recently arisen for electronic pulse circuits having response times in the range of approximately one to iifty millimicroseconds. Conventional circuits which use multi-electrode thermionic tubes cannot be effectively employed in this range primarily because the relatively high interelectrode capacitances of these tubes tend to round oi the sharp corners of any very narrow pulses present, delay their steep wave fronts slightly and broaden and reduce their amplitudes. Although the interelectrode capacitance can be reduced by modifications of the tube structure, these changes unfortunately are usually accompanied by a reduction in the tubes transconductance. Since the resolving time of a flip-hop pulse generating circuit, such as, for example, the multivibrator, is roughly proportional to the total capacitance to ground of a tube divided by its average transconductance, no appreciable net improvement in the response time is realized by redesigning the thermionic tube.

Although the phenomenon of secondary emission has been used successfully for a long time to amplify the minute photoemission current in photoelectric tubes, its application to thermionic tubes has lagged behind. The main reason for this lag is to be found in the difculty encountered in developing a coating with a high secondary emission ratio for the secondary cathode that does not evaporate or become contaminated at the relatively high temperatures existing in such tubes. The average number of electrons released from a surface by each incident or primary electron is known as the secondary-emission ratio. Certain combination surfaces, such as alkali halides on an alkali metal base and alkali oxides on various metal bases, have been developed recently which can withstand these high temperatures and give secondary emission ratios as high as eight per primary to eleven per primary, as contrasted with pure metals where the ratio is in the vicinity of one.

Due to the multiplying action of secondary emission, this type of tube has a maximum plate transconductance from four to eight times larger than what can be expected of regular pentodes. Also, a relatively small change in grid voltage can bring the tube from cutoff to saturation. These characteristics plus the fact that the ratio of the dynode current or the plate current to the associated electrode capacitance is very high for these tubes make them particularly advantageous in circuits where fast switching action is required.

Furthermore, since the grid voltage of a secondary emissive type tube is in phase with the dynode voltage, positive feedback action can simply be obtained by intercoupling dynode and grid together through a feedback network directly. In other words, no extra signal delay is added in this case by having to go through an extra tube for phase inversion. Thus, this tube makes it possible to have a feedback arrangement with the minimum amount of loop delay, thereby making ideal for high speed switching action and as a generator of short pulses.

Accordingly, it is a primary object of the present invention to provide an electronic pulse generator employing a secondary emissive tube in its control circuit for generating millimicrosecond pulses having extremely short rise times.

A secondary object of the present invention is to provide a discriminating circuit capable of rejecting rnillimicrosecond pulses whose amplitudes do not exceed a predetermined magnitude.

A further object of the present invention is to provide a pulse generator capable of producing pulses having a rise time of approximately six millimicroseconds and a variable width from approximately iifteen millimicroseconds and up at extremely high repetition rates.

A further object of the present invention is to provide a circuit capable of shaping millimicrosecond pulses.

A still further object of the present invention is to provide a counting circuit capable of registering millimicrosecond pulses and having `a resolving time of approximately twenty millimicroseconds.

Other objects and many of the attendant advantages of this invention will be readily appreciated as the same becomes better understood by reference to the following detailed description when considered in connection with the accompanying drawings, wherein:

Fig. 1 illustrates an electronic pulse circuit capable of performing as either a millimicrosecond pulse amplitude discriminator or a millimicrosecond pulse generating circuit; and

Fig. 2 illustrates a circuit arrangement which can be employed to count millimicrosecond pulses such as those produced by the above circuit.

Referring now to Fig. 1, which shows a preferred embodiment of the invention, and taking the case where the circuit is to perform as a millimicrosecond pulse generator, input pulses of suitable amplitude and rise time, such as, for example, those produced by a freerunning multivibrator, are coupled to the control grid of a secondary emission type amplifying tube 5. A differentiating network consisting of a series capacitor l and a shunt resistor 2 is included in the input circuit of this amplifying tube to prevent signals having a slow decay from causing spurious operation of the circuit. Amplitying tube 5 is normally maintained nonconducting by a negative bias applied to its control grid via the moving Contact of potentiometer 3 connected between a negative voltage source 26 and ground. The amplitude of the trigger signal is selected so that the positive trigger pulse obtained by diierentiating the leading edge of the pulse has suliicient amplitude to overcome the blocking bias and drive tube 5 to conduction. To improve the wave form of the negative pulse thus developed at the plate of tube S, a peaking inductance 7 is inserted in series with load resistor 6.

At the same time that this negative pulse occurs at the plate, a similar positive pulse is generated at dynode 4. This is due to the fact that upon electron bombardment the dynode loses more electrons than it receives from the cathode due to secondary emission. This positive pulse is directly coupled via capacitor 18 to the control electrode of a secondary emissive tube 20 lwhich is also normally maintained nonconducting by the negative voltage applied to its control grid from the juncture of series resistors i3 and 14 connected between the negative voltage source 26 and ground. By coupling the dynode 25 to one side of the capacitor 18, positive feedback is obtained and this tube performs as a monostable multivibrator having a stable and an unstable operating point, the tirst at cutot and the other at saturation, and an additional unstable operating point at an intermediate level of tube conduction. These various operating points are determined primarily by the dynode current versus dynode voltage characteristic of tube 20 and the magnitude of the parallel connection of dynode resistor 1G and the grid resistors 17 `and 15.

It is well known that if such a multivibrator rests at its permanently stable operating point` an instantaneous transition to the other stable operating point can only take place when the applied trigger signal, which in this case is the positive pulse obtained from dynode 4, has suiiicient magnitude and width to slide the circuit past the unstable point. If the amplitude of the trigger signal is insuicient, the tube returns to the permanently stable operating point from whence it initially started.

In the present case, the positive pulse at dynode 4 has suiiicient amplitude to bring about this transition and, as a consequence, tube 2t) is rapidly driven to saturation by the regenerative action of capacitor 18 and the leading edge of a negative output pulse appears in the plate circuit of this tube at terminal 24.

To increase the trigger sensitivity of the pulse generator further, the negative signal appearing at the plate of tube 5 is also coupled via capacitor 9 to the cathode of tube 20. This cathode is connected to `ground through a diode 19 in order to make the impedance between the cathode and ground large when the trigger signals are applied and small when tube 20 becomes conducting. In this way very rapid and precise triggering of tube 20 is obtained. Once tube 20 is driven to its temporary stable operating point, the circuit will stay in the vicinity of this point until the condenser 1S has accumulated enough charge so that the loop gain of the circuit becomes equal to unity. A jump back to a point beyond the permanent- `ly stable operating point now takes place after which the circuit again comes to rest. This process results in a positive pulse shaped behavior of the voltage at the dynode and a negative one at the plate. The time it takes for the circuit to go from the temporary stable operating point to the point of unity loop gain determines solely the width of the output pulse. This time, in turn, is determined by the magnitude of resistors 10, and 17, condenser 18, the grid bias, the location of the permanent and temporary stable operating points and the unity loop gain point. Therefore, the pulse width may be varied by making any of these quantities variable. As a practical point, however, it is usually easier to vary the grid-resistor 13, condenser 18 and the grid bias. Since varying the grid-resistor 15 and condenser 18 gives better linearity, these are usually used as the main pulse width control while the grid bias is merely varied for adjustment purposes.

In the case where the circuit of Fig. l is to perform as a millimicrosecond pulse amplitude discriminator, the discriminating level is set by adjusting the location of the movable tap of potentiometer 3. Only those input pulses whose amplitudes exceed the magnitude of the blocking bias thus supplied to the control grid of amplifying tube 5 will succeed in reproducing corresponding negative pulses at terminal 24.

Fig. 2 illustrates a counting circuit for registering a sequence of millirnicrosecond pulses during an interval of time up to a hundred microseconds. In other words, it is a kind of fast memory circuit that is able to store information coming in at a fast rate for a certain interval of time after which the information stored is destroyed. To prevent misiire due to input signals having different wave forms, it should be driven by a discriminator circuit of the type shown in Fig. l. When a positive pulse is applied to the grid of tube 33, which is normally cut off by virtue of the negative bias on its grid,

a positive pulse appears at the dynode of this tube and a negative one at its plate. These two electrodes are connected to the dynode of another secondary emission tube 38 via the diodes 36 and 35 respectively. These two diodes receive a bias from the attenuators consisting of resistors 46, 48, 49 and 52, -51, 53 in such a way that diode 36 is conducting when tube 38 is cut off and diode 35 is lconducting when tube 38 is conducting. The tube 38 is normally nonconducting by virtue of a negative bias on its grid supplied by the resistor 43 which is connected to the negative voltage supply 44. Also, it has a condenser connected between its dynode and its grid so that regenerative action may be created in the tube in such a way that it can temporarily have two stable operating points separated by an unstable one. Hence, if a positive pulse is initially applied to the input of the circuit, a positive pulse will appear on the dynode of tube 38 since diode 36 is conducting and diode 35 initially cut olf. This pulse will now turn tube 38 on and if no other pulses are applied to the circuit tube 38 will remain conducting for a time determined mainly by the size of the condenser 37 and resistor 54 and 53. When tube 3S becomes conducting, the bias on diodes 35 and 36 will now be such that diode 35 becomes conducting and diode 36 cut oli'. Hence, if a second pulse is applied to the circuit, the negative pulse at the plate of tube 33 will now pass over to the dynode of 38. Since this pulse is negative, it will then turn the tube off. Therefore, if the size of condenser 37, resistors 54 and 42 are made large enough so that tube 3S can remain conducting for a long time interval, tube 38 can be turned otr and on by sequences of fast pulses during this interval. Thus, tube 38 will act as a binary scaler being turned on and off by successive pulses applied to its input. In the plate of tube 38 a differentiating network is inserted consisting of essentially a condenser (41), a diode 60 and a resistor (40) connected up in such a way that when the tube initially turns on no pulse appears at its output while when the tube turns oi a positive pulse will appear. it will be understood, of course, that diode 69 acts as a clamp across load resistor 61 to insure the production of positive output pulses only. Thus, for a succession of pulses applied to the input of the circuit a positive pulse will therefore appear at the output for each second pulse applied and a scale o 2 is achieved.

Obviously many modications and variations of the present invention are possible in the light of the above teachings. It is therefore to be understood that within the scope of the appended claims the invention may be practiced otherwise than as speciiically described.

What is claimed is:

1. A millimicrosecond pulse generating circuit comprising, in combination, a secondary-emissive tube having as components thereof at least a cathode, control grid, dynode and anode, a diode connected between said cathode and a reference potential, means for biasing said control grid negatively with respect to said reference potential whereby said tube is normally maintained nonconductive, means -for coupling said anode via a load resistor to a voltage source positive with respect to said reference potential, a coupling capacitor interconnected between the dynode and control grid of said tube for providing a positive regenerative action therebetween, and means for simultaneously triggering said tube by applying a positive pulse and a negative pulse to the control grid and cathode, respectively, of said tube, said trigger pulses cooperating to move the operating point of said tube from its first stable point at tube cutoil" through its intermediate unstable operating point to its second stable operating point at plate current saturation whereby a negative going output pulse having an extremely fast lrise time is Vgenerated at the anode of said tube, and means for varying the time constant of the discharge path of said coupling capacitor thereby to determine the time at which said tube returns to its first stable .5 operating point and the trailing edge of said output pulse.

2. A millimicrosecond pulse generating circuit comprising, in combination, a secondary electron emissive vacuum tube having at least a cathode, a control grid, a dynode and an anode, a coupling capacitor connected between the dynode and the control grid of said tube so as to provide an extremely short time constant positive feedback path therebetween, means for connecting said control grid to a negative potential so as to normally maintain said tube nonconducting, a load resistor connected between said anode and a positive voltage, a diode, said diode being connected to said cathode and a source of reference potential and poled to pass positive pulse from said cathode to said reference potential, means for coupling simultaneously a positive trigger pulse to the control grid of said tube via said coupling capacitor and a negative trigger pulse to the ungrounded side of said diode whereby said tube is immediately driven to a condition of plate current saturation and whereby the leading edge of a negative output pulse is developed at the anode of said tube and means for controlling the time constant of the discharge path of said capacitor so as to establish the time at which said tube returns to its original nonconducting state and the time at which the trailing edge of said output pulse occurs.

3. A millimicrosecond pulse generator comprising, in combination, a iirst secondary-emissive tube having as components thereof a control grid, a dynode, a cathode and an anode, means yfor applying a blocking bias of a predetermined magnitude to the control grid of said tube for maintaining said tube normally nonconducting, means for applying positive input trigger pulses to the control grid of said rst tube thereby to render said tube conducting whenever the amplitude of said trigger pulses exceeds said blocking bias, said rst tube developing coincident positive and negative pulses at its dynode and anode, respectively, whenever it is rendered conducting, a second secondary-emissive tube, said second tube having as components thereof a cathode, a control grid, a dynode and an anode, means for normally maintaining said second tube nonconducting, a load resistor connected between the anode of said second tube and a positive voltage, a diode, said diode being connected between the cathode of said second tube and a reference potential, said `diode being poled to pass only positive pulses from said cathode to said reference potential, means for interconnecting the anode of said rst tube to the cathode of said second tube via a coupling capacitor, means for directly interconnecting the dynodes of said first and second tubes and a capacitor coupled between the dynode of said second tube and the control grid of said second tube, said capacitor providing a short time constant positive feedback path for said second tube whereby said second tube performs as a monostable pulse generator when simultaneously triggered by the coincident positive and negative pulses coupled to its control grid and cathode, respectively, from said first tube generates a negative going output pulse across said load resistor.

4. In a pulse generator as dened in claim 3, means for adjusting the discharge time constant of the capacitor coupled between the dynode of said second tube and the control grid of said second tube, thereby to regulate the duration of said negative going output pulse.

References Cited in the file of this patent UNITED STATES PATENTS 2,293,177 Skellett Aug. 18, 1942 2,294,782 Jacobsen Sept. 1, 1942 2,369,631 Zanarini Feb. 13, 1945 2,509,998 Mark May 30, 1950 2,597,796 Hindall May 20, 1952 2,847,565 Clapper Aug. l2, 1958 FOREIGN PATENTS 221,125 Switzerland Aug. 1, 1942

Images