US2365583A

US2365583A - Frequency-dividing circuits

Info

Publication number: US2365583A
Application number: US462998A
Authority: US
Inventors: George W Nagel; Mortimer A Schultz
Original assignee: Westinghouse Electric and Manufacturing Co
Current assignee: CBS Corp
Priority date: 1942-10-22
Filing date: 1942-10-22
Publication date: 1944-12-19
Anticipated expiration: 1961-12-19
Also published as: GB568556A; BE472352A; NL66092C

Description

Dec. 19, 1944. 5, w NAGEL ET AL 2,365,583

FREQUENGY-DIVIDING CIRCUIT Filed Oct. 22, 1942 Volaye B h n n n n nnflnnnnnnnnnnflnnnnnnflhlnnnnnn" M M 44-16 I .1 v 29 32 :9 I INVENTORS.

WITNESSES:

George 4 Na 1 and Mffld/ Mom? .5a ulfz,

- AT ORNE Patented Dec. 19, 1944 Q FREQUENCY-DIVIDING CIRCUITS George W. Nagel, Catonsville, and Mortimer A.

Schultz, Baltimore, Md., assignors to Westinghouse Electric & Manufacturing Company,

East Pittsburgh, Pa. vania a corporation of Pennsyl- Application October 22, 1942, Serial No. 462,998

2 Claims.

Our invention relates to frequency sub-dividers and, in particular, relates to arrangements in which multi-vibrator tubes are employed to derive an output voltage which has a frequency that is a sub-multiple of the frequency of an input voltage.

. One object of our invention is to provide a circuit in which an output voltage is derived which has a frequency which is a constant sub-multiple of the frequency of an input voltage.

Another object of our invention is to provide a circuit in which an output voltage is derived which is a very small constant sub-multiple of the frequency of an input voltage.

Still another object of our invention is to provide a circuit in which the ratio between the frequency of an output voltage and the frequency of an input voltage remains highly constant in spite of considerable variations in the electrical parameters of the circuit, such, for example, as variation in the plate circuit supply voltage, environing temperature, cathode emissivity of tubes used and the like. v

Qne of the more specific objects of our invention is to provide a circuit embodying a multivibrator in which an output voltage is derived, the frequency of which is a small and constant sub-multiple of the frequency of a voltage impressed on the input terminals of the system.

Other objects of our invention will become apparent upon reading the following specification, taken in connection with the drawing in which:

Figure 1 is a plot of certain voltage-againsttime curves useful in explaining our invention; and

Fig. 2 is a schematic diagram of one form of circuit embodying our invention. v

It is desirable for numerous purposes in the radio field to be able to derive a periodic voltage, the frequency of which is a constant, and often a very small, sub-multiple of the frequency of a periodic input voltage. One arrangement for effecting this result, which has previously been used, has been to provide a. multi-vibrator circuit embodying a pair of tubes arranged so that current flows at the beginning of their periodiccycle through one tube, while the other remains nonconductive; and that in the course oftime, the charging of a condenser causes the first tube to lose its conductivity and current flow to be initiated through the second tube. Usually current flow through the second tube is arranged, by proper choice of electrical circuit constants, to continue for only a comparatively short time, whereupon the second tube becomes non-conductive and the first tube becomes again conductive, thus returning the system toits original condition and completing a periodic cycle.

The tubes I and 2 of Fig. 2 herein are interconnected in one type of circuit capable of causing such a behavior as has just been described and constitute a cmulti-vibrator. An input circuit impresses voltage of the-higher frequency of which it is desired to derive a sub-multiple between the control electrode of the tube and ground, and currents of the sub-multiple frequency may be derived from the resistor 9.

We have found that the simple employment of the multi-vibrator just described results in an arrangement which is liable to disturbance from spuriou and accidental voltage fluctuations in the input circuit waves, and from inductive effects due to neighboring circuits and apparatus. We have also found that the difficulty from such spurious effects just described can be reduced if a buffer amplifier, that is to say, an ordinary amplifier tube, is inserted in the input circuit ahead of the connection to the grid of the tube I. With such an arrangement, if a periodic voltage of high constant amplitude is impressed on the input circuit of th amplifier, it is possible to derive an output voltage which will remain in frequency uniformly at one-tenth of the frequency at the input pulses; but if the attempt is made to divide the frequency in a ratio much greater than this value of 1'0 to 1, reliability of operation cannot be assured in practical service. Variations of line voltage, environing temperature, cathode emission and the like will result in variations of the ratio of frequencie if the attempt is made to make it substantially greater than 10 to 1.

Referrin to Fig. 1, we have found that after the tube 2 in Fig. 2 (which tube we assume to be the one remaining non-conductive during the earlier portion of the output cycle and to maintain conductivity for only a relatively short time during the end portion of that cycle) has thus fired and become non-conductive, the voltage required to reinitiate current flow in it follows a l curve similar to AA in Fig. 1; in other words,

the voltage required to transfer current flow to the tube 2 in Fig. 2 from the tube I quickly rises to a high value at the beginning of the cycle of output current and then falls along a curve of gradually decreasing slope. The curves B-B in Fig. 1 represent the effects of the pulses or waves impressed on the grid 5 of tube l in Fig. 2, and, if of constant magnitude, will sooner or later intersect the curve A-A of Fig. 1. When this intersection occurs, current flow is transferred from the tube l to the tube 2 of Fig. 2 and the output wave passes through its brief cycle phase of voltage decrease to end the cycle.

The action just described illustrates the behavior of a multi-vibrator having pulses of constant magnitude impressed on its input electrode;

and since the slope of the curve AA is concurve AA will change the number of the pulses B which elapse before the crest of the output wave is produced by transfer of current flow from the first to the second tube of the multi-vibrator. That is to say, if in the example taken it is attempted to derive an output wave which has a period equal to 50 pulses of the input wave; reliability of such frequency division cannot be relied upon, since a slight change in the electrical parameters of the circuit is liable to cause the output curve to continue through only 49 or perhaps 51 pulses of the input wave.

In accordance with our invention, we arrange that the input wave, instead of being of constant amplitude, shall be of relatively small amplitude during the earlier portions of the output cycle, but shall rise sharply in amplitude as the intended final pulse of the input wave is approached. This is illustrated by the form of the curves B-B in Fig. 1 which clearly shows the increased certainty that intersection of the curve AA will occur at the 50th pulse rather than at the 49th pulse in the example taken. Any slight variation in circuit parameters which causes minor increases or decreases in ordinates of the curves AA or the curves BB will have little effect in preventing intersection of the curve AA with pulse No. 50.

Fig. 2 shows details of a circuit adapted to realize the conditions illustrated in Fig. 1., Thus I and 2 represent a pair of electrical discharge tubes which are preferably of the high-vacuum type and incorporate electron-emissive cathodes 3-4, control electrodes 5-6 and anodes Ia-B. The cathodes 3-4 are connected to each other and to one terminal of a resistor 9, the other terminal of which is grounded. The anode I is connected through a capacitor I0 to the grid 6 of the tube 2 and also through resistor 32 to a source of potential I5 which makes it positive with respect to ground. Grid 6 is also connected to one terminal of a resistor I4, the other terminal of which is connected to a tap on a source I2 of bias potential. A capacitor I3 is preferably connected between the tap on the potentiometer I2 and the positive terminal thereof which is grounded. The control electrode 5 of the tube I is connected to ground through a resistor II.

The tubes I and 2 may be considered as comprising the multi-vibrator portion of the circuit.

An amplifier tube I6, which preferably is of the high-vacuum type and comprises a cathode H, a control grid I8, and may also be provided if desired with a screen grid I9 and a suppressor grid 20, has its anode 2| connected through a capacitor 22 t0 the control electrode 5 of the tube I. The anode 2| is connected through a resistor 23 to the positive terminal of a directcurrent source 24, the negative terminal of which is grounded. The cathode I! of the tube I6 is connected through a capacitor 25 and, if desired, through a resistor 26 to ground. The input circuit 21 is connected to impress voltage across the resistor 26. A potentiometer 28 is connected to impress a variable bias voltage relative to ground upon the cathode II. The positive terminal of the direct-current source 24 is connected through a resistor 29 to the control grid I8 of the tube I6, and the latter is connected through a capaci tor 3I to ground. The grid 20 is connected to cathode I1 and grid IE! to the positive terminal or source 24. The grid I8 of tube I6 is likewise connected to the anode 8 of the tube 2. The output voltage of the system may be tapped off the resistor 9.

The operation of the above-described arrangement is as follows: Starting with the capacitor 3| in a nearly discharged condition, the tube 2 will be non-conductive since its control electrode 6 is biased negatively by the position of the tap on the potentiometer I2, while its cathode is positive relative to the positive terminal of this potentiometer because of current flow in the tube I through the resistor 9. The tube I is itself conducting since, by selection, it is of a type which, in absence of any such negative bias as the potentiometer I2 impresses upon the control electrode 6 of tube 2, conducts current readily when its control electrode 5 is at ground potential, notwithstanding the fact that its cathode 3 is biased slightly positive relative to ground by its current flowing through the resistor 9. This current through the tube I is derived from the direct-current source I5 and is limited in magnitude by the resistor 32 and the characteristics of tube I.

The tube I6 is non-conductive, notwithstanding the impression of positive voltage on its anode 2I by the voltage source 24, since it is, by selection, chosen of such a type that, when the capacitor 3| is (as previously stated) nearly discharged, the positive potential impressed by the potentiometer 28 on cathode II relative to ground will prevent current flow in tube I6.

- The negative input voltage pulses of the higher frequency are impressed across the resistor 26, and therefore effectively between grid I8 and cathode I1 of tube I6, but under the cihcucmstances just described, they cause no current flow through the tube I6, and hence the coupling of the anode 2I thereof through capacitor 22 to the control electrode 5 of tube I produces little or no voltage pulse on the last-mentioned control electrode relative to its cathode 3.

In the course of time, the resistor 3I will gradually charge up from source 24 through resistor 29 and make the control electrode I8 of the tube I6 more positive. A' condition will eventually be reached, therefore, when the potential of the control electrode I8 is high enough so that current will flow through the tube I5, and as the input Voltage pulses across resistor 26 cause corresponding fluctuations in potential between the cathode I1 and the control electrode I8, current pulses of the input frequency will flow from the source 24 through the resistor 23 and anode of tube I6, to cathode I'I, through capacitor 25 and resistor 26 to ground. These voltage pulses of the input frequency and of rapidl increasing magnitude will pass through the capacitor 22 and thereby impress themselves as negative pulses on the control electrode 5 of the tube I. They will, accordtube I, through resistor 9 to ground. These pulses of the input frequency through resistor 9 impress themselves as negative pulses making the cathode 4 of tube 2 less positive relative to its control electrode 6, Since they become of substantial magnitude directly after the initiation of current flow through the tube I6, they act as a powerful agent to immediately render tube 2 conductive and to permitdischarge of capacitor 3| through the anode 8 thereof to its cathode 4 and thence through resistor 9 to ground. The strong 7 positive pulses thus impressed by this current fl-ow from tube 2 on resistor 9, taken in conjunction with the negative pulses already described as being impressed on control electrode 5 through capacitor 22, simultaneously rendering the tube I non-conductive. I

The resistor 9 and internal resistance of tube '2 are made to be of such values that capacitor 3| quickly discharges itself through tube 2, thereby cutting off current flow through tube I6 and at the same time rapidly reducing the current through tube 2 by the reduction of voltage impressed by capacitor 3| on its anode 8. Cessation of current flow through tube l6 cuts off the negative pulses flowing from anode 2| through capacitor 2'2 and resistor I4 to the control grid of tube I and permits the latter to return to a conductive condition as soon as the current flow through tube 2 has fallen to a point where the resultant bias across resistor 9 is insufficient in itself to keep tube Icut 01f. The entire system has, ac-

cordingly, returned to the original condition initially assumed in which the capacitor 3| is sub stantially discharged, the tube I conducting and the tubes 2 and I6 non-conducting. In consequence, the above-mentioned cycle can start all over again and repeat itself indefinitely.

It will be seen from the foregoing that the pulses of higher frequency Voltage impressed from the input circuit upon tube I are of very nearzero magnitude until capacitor 3| has charged up high enough to render tube I6 conductive; and that they then rapidly rise to much higher values in the manner illustrated by the pulses B-B m Fig. 1.

Values which will be readily evident to those skilled in the art may obviously be assigned to the various elements of the above-described circuit, so that the setting of the variable tap on potentiometer I2 and the values of capacitor 3| and resistor 29 will be predominant in, and virtually control, the time at which tube I6 becomes conductive, and so will determine the number of pulses of the higher frequency input voltage from circuit 21 which will elapse before current flow shifts from tube to tube 2. Thus minor variations inthe voltage of sources 32 and 24 and in the internal characteristics of tubes 2 and I6 will havelittle eifect in determining the number of cycles of input source 21 which will elapse before one cycle of the output voltage across resistor 9 occurs. In other words, the values of the resistor 29 and the capacitor 3| and the bias setting of the potentiometer I2 will bepredominant over any ordinary fluctuations met with in other circuit parameters in determining the ratio of sub-division of the frequency of the input voltage 21 which is obtained by this system.

We claim as our invention:

1. In combination, a first tube having a cathode, anode and control electrode, a second tube having a cathode, anode and control electrode, the two cathodes being connected together to one terminal of a resistor having its other terminal connected to ground, a, connection through an output resistor from the positive terminal of a directcurrent source of which the negative terminal is groundedto the anode of said first tube, means for supplying current to the anode of said second tube, 'a connection through a capacitor from the positive terminal of said first tube to the control electrode of said second tube, means for impressing a negative bias relative to ground on the lastmentioned control electrode, a third tube having a cathode, at least one control electrode and an anode, a connection through a capacitor and an input resistor from the last-mentioned cathode to ground, means for impressing a positive bias potential relative to ground on the last-mentioned cathode, a source of direct-current voltage having its negative terminal grounded and its positive terminal connected through an input resistor to the anode of said third tube, a connection between the last-mentioned anode through a capacitor to the control electrode of said first tube and a connection from the last-mentioned controlode, anode and control electrode, a second tube having a cathode, anode and control electrode, the two cathodes being connected together to one terminal of a resistor'having its other terminal connected to ground, a connection through an output resistor from .the positive terminal of a direct-current source of which the negative terminal is grounded to the anode of said first tube,

means for supplying current to the anode of said second tube, a connection through a capacitor from the positive terminal-of said first tube to the control electrode of said second tube, means for impressing a negative bias relative to ground on the last-mentioned control electrode, a third tube having a cathode, at least one control electrode and an anode, a connection through a capacitor and an input resistor from the last-mentioned cathode to ground, means for impressing a positive bias potential relative to ground on the lastmentioned cathode, a source of direct-current voltage having its negative terminal grounded and its positive terminal connected through an input resistor to the anode of said third tube, a connection between the last-mentioned anode through a capacitor to the control electrode of said first tube and a connection from the last mentioned control electrode through a, resistor to ground; and means for impressing a voltage which increases with time relative to ground potential on the control electrode of said third tube.

GEORGE W. NAGEL. MORTIMER A. SCHULTZ.