US2529547A - Frequency divider - Google Patents

Description

Nov., 14, 1950 J. F. FISHER FREQUENCY DIVIDER Filed March 23, 1948 IN VEN TOR. www f n n/m TIME Patented Nov. 14, 1950 FREQUENCY DIVIDER Joseph F. Fisher, Philadelphia, Pa., assignor to Philco Corporation, Philadelphia, Pa., a corporation of Pennsylvania Application March 23, 1948, Serial No. 16,410

(Cl. Z50-27) 8 Claims.

The invention herein described and claimed relates to an improved frequency-dividing circuit of the counting type. The improved circuit is adapted to achieve frequency division, of high count-down ratio, with excellent stability. For example, the circuit is capable of accomplishing frequency division of excellent stability with a divisor as large as twenty or more. This is in contrast to conventional counting circuits which are ordinarily unable to divide by a factor larger than six or eight without losing stability.

The operation of the improved circuit of the present invention is similar in some respects to that of prior art circuits in that frequency division is effected by cumulating a voltage in incremental fashion and utilizing a cumulative voltage of predetermined magnitude to trigger a blocking tube oscillator, thus to produce pulses at a sub-frequency of the frequency of the applied pulses,

The improved circuit of the present invention is quite dierent, however, from that of prior art circuits in that the new circuit includes means whichenable the circuit to produce, from recurrent applied pulses of substantially equal amplitude, voltage increments of increasing amplitude. This is in contrast to most prior art circuits which derive, from recurrent applied pulses of equal amplitude, voltage increments of decreasing amplitude.

Stated graphically, the present invention provides a counting circuit capable of producing a voltage staircase whose risers increase in amplitude. Thus, in the preferred manner of operation, the riser which effects firing of the blocking oscillator tube, is larger in amplitude than the rst step of the staircase. Positive and dependable flring action is thus assured. Most prior art circuits, on the other hand, produce a voltage staircase whose risers decrease in amplitude; hence, the amplitude of the riser intended to eifect firing may be too small to assure dependable counting action.

If desired, the new circuit may be so operated that the lower steps of the voltage staircase diminish in amplitude in customary manner, reserving the increase in riser amplitudes to the higher steps, particularly the steps immediately prior to, and including, the trigger step which is intended to fire the blocking oscillator. Or, if desired, the new circuit may be so adjusted as to produce a voltage staircase whose risers are substantially equal in amplitude.

While I know of several other frequency-dividing circuits of the counting type capable of producing voltage staircases whose risers are of substantially constant amplitude, I know of no other frequency-dividing circuit capable of producing a voltage staircase whose risers increase in amplitude. It will be apparent to those skilled in the art that -my new circuit is capable of achieving very accurate and stable frequency division since the riser intended to effect firing may be made of adequate amplitude to assure dependable action.

It is an object of this invention to provide a frequency-dividing circuit of the counting type capable of effecting accurate division with divisors as large as twenty or more.

It is another object of this invention to provide a frequency-dividing circuit adapted to develop, from successively applied voltager pulses of substantially equal amplitude, voltage increments of increasing magnitude, whereby the trigger step of the voltage staircase may be larger than the first step.

Another object of this invention is to provide, in a frequency-dividing circuit of the counting type, means for compensating, and, in the preferred manner of operation, over-compensating, for the exponential decrease in the amplitude of the incremental voltages which ordinarily characterizes conventional counting circuits.

These and other objects, advantages and features of the present invention, and the manner in which the objects are attained, will become clear from a consideration of the following detailed description and of the accompanying drawing wherein:

Figure 1 is a schematic representation of a preferred form of my improved frequency-dividing circuit; and

Figures 2 and 3 are graphical representations which will be helpful in describing and understanding the present invention.

Referring now to Figure 1, there is shown a frequency-dividing circuit of the counting type comprising a source il) of voltage pulses, a slightly modied but otherwise conventional step-by-step counting circuit l l, a blocking tube oscillator circuit l2, and, in accordance with the present invention, a compensating or booster circuit i4,

Source IG may conveniently comprise a cath--l ode-follower driver circuit represented generally at i3 which delivers positive voltage pulses of substantially equal amplitude through an internal resistance l5.

Circuit Il comprises a coupling capacitor I6, a charge diode Il, a discharge diode i3, anda storage capacitor I9. Storage capacitor I9 will be quite large, ordinarily, relative to coupling capacitor I6. For example, storage capacitor I9 may be fty times larger than coupling capacitor I6. The cathode of charge diode I'I is connected in the manner shown in the drawing, to the upper plate of storage capacitor I9, to the grid of booster tube 2|, and, by way of winding 22 of transformer 23, to the grid '24 of blocking oscillator tube 25. The anode of charge diode I1 is connected to the cathode of discharge diode I8 and to the low potential plate of coupling capacitor I6. The anode of discharge diode I8 is connected to a source of negative direct-current potential, B--.

Circuit I2 may conveniently comprise a known form of blocking tube oscillator including a triode 25 and a pulse transformer 23 whose primary winding 26 is connected between the anode 21 of the triode 25 and a source of positive plate potential, B+. The secondary winding 22 of pulse transformer 23 is connected, as previously indicated, between the grid 24 of tube 25 and the upper plate of storage capacitor I9. The cathode 28 of tube 25 may be connected directly to ground. The sub-multiple frequency output of the counting circuit may be taken between the plate 21 of oscilator tube 25 and ground, as shown in Figure 1.

Compensating or booster circuit I4 comprises a tube 2| Whose cathode 29 is connected, by way of a load resistor 3D, to a source of negative direct-current potential, B-. The potential of the supply voltage, B-, is higher than that of the supply voltage, B--, connected to the anode of diode I8. The voltage difference between B- and B-- constitutes the xed bias on booster tube 2I, as will` become clear. Storage capacitor I9 is connected directly between cathode 29` and grid 2i) of tube 2l. Plate 3l of tube 2I is connected to a source of positive plate voltage, B+.

It will be seen that, except for the compensating or booster circuit I4, the circuit of the present invention is similar to the conventional prior art counting circuit. In the prior art circuit, the low potential plate of storage capacitor I9 is connected directly to ground.

Referring now to Figures 2 and 3, there is shown, in Figure 2, a graphical representation of the charge developed in step-by-step fashion across storage capacitor I9, and, in Figure 3, there is shown a graphical representation of the staircase voltage developed in step-by-step fashion between the grid 24 and cathode 28 of oscillator-tube 25.

The operation of the circuit will now be described. Assume that blocking oscillator tube 25V has fired and that storage capacitor I9 has just been discharged through resistor 30 by the grid current of oscillator tube 25. Assume that tube 25 has now been cut off, and that the circuit is. awaiting the arrival of the next pulse. At this stage in the operation of the circuit, the potential of grid 24 of oscillator tube 25, and of the upper plate of capacitor I9 (point d in the circuit of Figure l), is at a value which corresponds to the voltage at the bottom of the staircases depicted graphically in Figures 2 and 3. It will be seen that the value of the voltage at the bottom of the staircase is determined by ther value of negative supply voltage, B--, connected to the anode of diode I8. For, when tube 25 res and,grid current flows to discharge capacitor I9, the upper plate of capacitor I9 is driven down to a potential limited by, and hence substantially equal to, the voltage of the nega- 4 tive supply, B--. For example, the supply voltage, B--, may be minus nity-five volts, as

is indicated in Figure 1. In such case, the voltage at the bottom of the staircase of Figures 2 and 3 is substantially minus fifty-five volts. This voltage, i.e., the voltage of the negative supply, B--, appearing on grid 24 and on the upper plate of capacitor I9 when tube 25 cuts off, also appears on grid 29 of booster tube 2i.

The potential of cathode 29 of tube 2I (point e in Figure 1), prior to the arrival of the first pulse following discharge, is higher than that of point d, the potential of point e being equal to that of the negative supply voltage, B-, which, by way of example, may be minus fifteen volts. Hence,l the fixed bias on tube 2I is the voltage difference between the negative supply voltages, B-- and B-, which, in the present example, is minus forty volts. This voltage appears, then, acrossv capacitor I9 prior to the arrival of the first pulse following discharge.

When the leading edge of the rst pulse following discharge arrives, the potentials on both sides of coupling capacitor I6 rise by an amount equal to the amplitude of the pulse. Diode II thereupon conducts, but diode I8 remains nonconductive. During conduction, diode I'I comprises a virtual short circuit; its relative resistance is extremely low, and the voltage drop thereacross is negligible. The potentials of both sides of capacitor I9 rise by an amount equal substantially to the amplitude of the applied pulse, and the pulse voltage is impressed across resistor 3D. Resistor 30 is of relatively small value, and capacitors I6 and I9 quickly charge, in inverse proportion to their respective capacitances. The R. C. time constant of the charging circuit is substantially shorter than the duration of the applied pulse. For example, in a typical case, the R. C. time constant may be of the order of one microsecond, and the pulse duration may be ve microseconds. Capacitors I6 and I9 therefore charge to their full extent prior to the termination of the pulse. The values of capacitors I6 and I 9 are so selected, with respect to each other, that a desired division of the applied pulse voltage is eifected therebetween, and a desired proportion of the voltage of each pulse is developed across storage capacitor I9.

When the trailing edge of the` rst pulse arrives, both sides of capacitor I6 suddenly drop in potential by an amount substantially equal to the magnitude of the pulse. The cathode of diode I8 is then at a lower-potential than the anode. Diode I8 thereupon conducts, while diode I1 ceases to conduct. Capacitor I6 discharges through diode I8 until the potential of point c rises to the potential of the supply voltage, B--, which, in the present example, has been assumed to be minus fifty-five volts. The potential of the cathode of diode I'I is now higher than that of the anode, i. e, the potential of point d is now higher than that of point c by the amount of charge developed across capacitor I9 by the first pulse. This charge is held on capacitor I9 for there is no discharge path to the lower plate of capacitor I9 (diode I'I, tube 2I, and tube 25 all being in non-conductive condition). The charge developed across capacitorv I9by the rst pulse is represented in Figures 2` and 3 by the rst riser. The potential of the cathode of tube 2'I, point e in the circuit of Figure l, remains equal to that of the supply voltage, B-, so long as tube 2I remains in a non-conductive state. In the present example, point e is minus fifteen volts, when tube 2I is non-conductive.

I have just described that, as a result of the first pulse, the potential of point d was increased while the potential of point e remained unchanged. Stated another way, the first pulse effected a decrease in the negative grid-to-cathode voltage on tube 2|, the magnitude of the decrease being equal to the incremental charge developed across capacitor I9. Assume that this decrease in negative polarity of grid 20 was insufcient to cause tube 2| to conduct, and that tube 2| remains non-conductive following the first pulse.

When the leading edge of the second pulse arrives, diode I'I does not conduct until point c has risen to the potential of point d. Thereupon diode I1 conducts. Diode I8, of course, remains nonconductive throughout the duration of the pulse. It will be seen that only the voltage difference, between the amplitude of the applied pulse and the charge across capacitor I9 derived from the previously applied pulse, is divided proportionately between capacitors I6 and I9. Hence, the second pulse is effective to develop across capacitor I9 a voltage increment whose magnitude is less than that developed thereacross by the first pulse. The second riser of the voltage staircases of Figure 2 and 3 is therefore shown to be smaller in amplitude tharn that of the rst riser.

The grid-to-cathode Voltage on tube 2| has now been made less negative by the sum of the charges developed across capacitor I 9 by the rst two pulses. Assume that this decrease in the negative polarity of grid 2i) is not sufficient to permit tube 2| to conduct. The potential of cathode 29 (point e) therefore remains unchanged.

The action described above continues, as is indicated in Figures 2 and 3 by the third, fourth and fth staircase risers of decreasing amplitude, until the potential of grid 29, with respect to that of cathode 29, is raised sufciently to cause tube 2| to conduct. When this occurs, plate current flows through load resistor 30, and the potential of cathode 29 and of the lower plate of capacitor I9 (point e in the circuit of Figure 1) becomes positive with respect to that of the supply voltage,

B-. In other words, the potential of the lower plate of capacitor I9, heretofore at the potential of the supply voltage, B-, rises by the amount of the voltage drop across load resistor 30. And, since no charging path exists, the upper plate of capacitor I 9 rises by an equal amount.

Observe then that the conduction of booster tube 2| causes no change in the grid-to-cathode voltage of the tube since no charge is developed across capacitor I9 as a result of the tube conducting. Note, however, that while the grid-tocathode voltage of booster tube 2| remains unchanged, the grid-to-cathode voltage of the blocking oscillator tube 25 does change when the booster tube 2| conducts, for the potential of point d has risen, with respect to ground, by the amount of voltage developed across load resistor 30.

Assume that, in the present example, booster tube 2| conducts for the first time when the sixth pulse is received. The incremental charge def I9 lby, the sixth pulse. The total rise in the gridto-cathode voltage of oscillator tube 25, as a. result of the arrival of the sixth pulse, is therefore the sum of two components. And, if the circuit constants be properly selected, this sum may be larger than the increase occasioned by the fth pulse. The sixth riser in Figure 3 is, therefore, shown to be larger than the fifth.

Following the termination of the pulse which caused booster tube 2| to conduct (the sixth pulse in the present illustration), tube 2| continues to conduct. As the succeeding pulses arrive, capacitor I9 is charged incrementally by each pulse, the magnitude of each increment being less than that of previous increments. This is shown graphically in Figure 2. The grid-tocathode voltage of booster tube 2| thus becomes less and less negative, or more and more positive, in step-by-step fashion in a manner corresponding to the staircase of Figure 2, and the plate current through load resistor 3|]` becomes larger and larger. As a result, the voltage drop across load resistor 3i! increases step by step and the potential of point e, and hence of point d, becomes more and more positive in step-by-step manner relative to the voltage source, B-.

Observe that the total change occurring in the grid-to-cathode voltage of blocking oscillator tube 25 as each pulse is received is equal to the change in voltage occurring between point d and ground. As described above, this change is comprised of two components, one of which is the incremental charge developed by each pulse across capacititor I9. The other component is the rise occasioned by each pulse in the potentials of points e and d, which as described above results from the rise in the grid-to-cathode voltage of tube 2| and the increase in plate current through load resistor 39. The magnitude of the latter component is a function of the selected circuit constants, including the gain of tube 2|. This component may be of such magnitude that, when added to the incremental charge developed across capacitor I9 by an applied pulse, the sum may constitute a voltage change as large as, or larger than, that of the rst riser of the staircase. This is shown graphically in Figure 3 which, as previously indicated, depicts the step-by-step change in the grid-to-cathode voltage of the blocking oscillator tube 25. In Figure 3, the point at which booster tube 2| begins to conduct is appropriately marked, as is the potential at which the blocking oscillator tube 25 fires.

It will be understood that tube 2| operates as an amplifier tube and not as a cathode follower, for the voltage developed in step-by-step fashion across capacitor I9 is applied directly between the grid 29 and cathode 29 of tube 2 I. Observe that the voltage is not applied between the grid 20 and the low potential end of the cathode load resistor 30, as would be the case in cathode-follower operation. By suitable selection of circuit constants, the gain of tube 2| may be made substantially as large as desired. The amount of gain desired will depend, among other things, upon the count-down ratio desired to be achieved. In general, the larger the frequency division to be effected, the smaller the gain desired.

It will be apparent to those skilled in the art that in some instances, as, for example, where the divisor is relatively small, the circuit parameters, particularly the fixed bias, may be so selected that tube 2| begins to conduct Vupon arrival of the first pulse of the cycle, i. e., upon arrival of the rst pulse following the firing of almaar? 7 oscillator tube 25 and the discharge of capacitor I9. In such case, all of the risers of the voltage staircase of Figure 3 may be of increasing amplitude.

If desired, all of the risers of the Voltage staircase may be made of substantially equal amplitude. This may be accomplished by so selecting the circuit parameters that the decreasing rate of rise in the voltage developed across capacitor I9 is substantially compensated by an increasing rate of rise in the voltage developed across load resistor 30. The change occurring in the potential of point d with respect to ground will then be linear. Ordinarily, however, it will be more desirable to develop a staircase whose slope or rate of rise increases.

It will be understood that such things as the slope of the staircase and the number of staircase steps required to effect ring of the blocking oscillator tube are functions of the circuit constants, including such parameters as the fixed bias on tube 2I, the plate-current grid-voltage characteristic or mutual conductance of tube 2 I, the value of load resistor 30, and the relative values of capacitors I6 and I9. For example, increasing the fixed bias of tube 2l, other things remaining unchanged, will increase the number of steps required to effect firing of blocking oscillator tube 25, whereas increasing the value of load resistor 3D will increase the rate of rise or slope of the staircase, and, other things remaining unchanged, will decrease the number of steps required to effect firing.

Indicated in Figure l are the circuit constants used in a particular frequency-divider circuit which I have built in accordance with the present invention and which has produced very satisfactory results. The constants shown are, of course, by way of illustration only and are not intended to be limiting.

Having described my invention, I claim:

l. A frequency-dividing circuit comprising: a blocking tube oscillator; a source of recurrent voltage pulses of substantially equal amplitude; means, including a storage capacitor, responsive to said recurrent voltage pulses for cumulating incremental voltages whose amplitudes successively decrease; means, including an amplifier tube, responsive to said cumulating incremental voltages for developing and cumulating other incremental voltages whose amplitudes successively increase; means for combining additively in an aiding sense both of said cumulating voltages; and means for applying said combined voltages to said blocking tube oscillator.

2. A frequency-dividing circuit comprising: a blocking tube oscillator; an amplifier tube having at least cathode, plate and grid electrodes and having a load impedance in its plate-to-cathode circuit; a voltage-divider circuit comprising a rst capacitor and a second capacitor; a source of recurrent voltage pulses of substantially equal amplitude; means for connecting said source across said voltage-divider circuit; means for effecting discharge of said first capacitor in the intervals between pulses while maintaining the charge on second capacitor, whereby a voltage cumulates in step-by-step manner across said second capacitor; neans for applying said cumulating voltage between the grid and cathode of said ampliiier tube to develop an output voltage across saidY load impedance, the amplitude of said output voltage being a function of said applied voltage; means for combining additively in an aidingv sense thev cumulating voltage de- 8 veloped across said second capacitor and the output voltage developed across said load impedance of said amplifier; and means for applying said combined voltage to said blocking tube oscillator.

3. A frequency-dividing circuit comprising: a blocking tube oscillator; an amplifier tube having at least cathode, plate and grid electrodes and having a load impedance in its plate-tocathode circuit; a voltage-divider circuit comprising a first capacitor and a second capacitor; a iirst diode; a second diode; a source of recurrent voltage pulses of substantially equal amplitude; means for connecting said source across said voltage-divider circuit; means for so connecting said first diode to said first capacitor as to effect discharge of said capacitor in the intervals between pulses; means for so connecting said second diode to said second capacitor as to oppose discharge of said second capacitor during the intervals between pulses, whereby a first staircase voltage develops in step-by-step fashion across said second capacitor; means for applying said staircase voltage between the grid and cathode of said amplifier tube to develop across said load impedance a second staircase voltage whose slope is of the same polarity as that of said first staircase voltage; means for combining additively said first and second staircase voltages and means for applying said combined voltages to said blocking tube oscillator.

4. A frequency-dividing circuit comprising: a blocking tube oscillator; an amplifier tube having at least cathode, plate and grid electrodes and having a load resistance in its plate-to-cathodc circuit; a voltage-divider circuit comprising a rst capacitor and a second capacitor; a first diode; a second diode; a source oi recurrent voltage pulses of subtantially equal amplitude; means for connecting said source across said voltage-divider circuit; means for so connecting said first diode to said first capacitor as to effect discharge of said capacitor in the intervals between pulses; means for so connecting said second diode to said second capacitor as to oppose discharge of said second capacitor during the intervals between pulses, whereby a rst staircase voltage develops in step-by-step fashion across said second capacitor, said first staircase voltage having a decreasing rate of rise; means for applying said rst staircase voltage between the grid and cathode of said ampliiier tube to develop across said load impedance a second staircase voltage whose slope is of the same polarity as that of said rst staircase voltage; means for combining additively said iirst and second staircase voltages and means for applying said combined voltages to said blocking tube oscillator.

5. A frequency-dividing circuit comprising: a blocking tube oscillator; a source of recurrent voltage pulses of substantially equal amplitude; a voltage-divider circuit comprising a coupling capacitor and a storage capacitor; means for applying said voltage pulses across said voltagedivider circuit; means, including a first rectiiier element connected between said coupling capacitor and said storage capacitor, for maintaining the voltage developed across said storage capacitor; means, including a second rectifier element connected between said coupling capacitor and a point of reference potential, for discharging said coupling capacitor to said reference potential in the intervals between pulses; an amplifier tube having at least cathode, plate and grid electrodes and having a cathode load impedance in the plate-to-cathode circuit; means for applying the voltage developed across said storage capacitor between the grid and cathode of said amplier tube to develop a voltage of the same polarity across the said cathode load impedance; means for combining additively the voltage developed across said cathode load impedance and the voltage developed across said storage capacitor; and means for applying said combined voltages to said blocking tube oscillator.

6. A frequency-dividing circuit comprising: a blocking tube oscillator; a source of recurrent voltage pulses oi substantially equal amplitude; a Voltage-divider circuit including a coupling capacitor, a storage capacitor and a rst rectifying element connected in series between said capacitors; means for cumulating across said storage capacitor in step-by-step fashion incremental voltages of decreasing amplitude, said cumulating means including a second rectifying element, one electrode of said second rectifying element being connected to the common junction of one side of said coupling capacitor and a diierent electrode of said Iirst rectifying element, the other electrode of said second rectifying element being connected to a point of low reference potential; an amplier tube having at least cathode, a plate and grid electrodes and having a cathode load impedance in the plateto-cathode circuit; means for applying the voltage developed across said storage capacitor between the grid and cathode of said amplifier tube; means for combining additively the voltage developed across said cathode load impedance of said amplifier tube and the voltage accumulated across said storage capacitor; and mens for applying said combined voltages to said blocking tube oscillator.

7. A frequency-dividing circuit comprising: a blocking tube oscillator; a source of recurrent voltage pulses of substantially equal amplitude; means, including a storage capacitor, responsive to said recurrent voltage pulses for developing a rst Voltage of stair-step Waveform; means, including an amplifier tube, responsive to said rst stair-step voltage for developing a second voltage of stair-step waveform Whose slope is or" the same polarity as that of said first stair-step voltage; means for combining additively said rst and second stair-step voltages; and means for applying said combined voltages to said blocking tube oscillator.

8. A frequency-dividing circuit comprising: a blocking tube oscillator; a source of recurrent voltage pulses of substantially equal amplitude; means, including a storage capacitor, responsive to said recurrent Voltage pulses for developing a first voltage of stair-step waveform; means responsive to said first stair-step voltage for developing a second voltage of stair-step waveform whose slope is of the same polarity as that of said rst stair-step voltage; means for combining additively said rst and second stair-step voltages; and means for applying said combined voltages to said blocking tube oscillator.

JOSEPH F. FISI-IER.

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

UNITED STATES PATENTS Number Name Date 2,113,011 White Apr. 5, 1938 2,114,016 Dimond Apr. 12, 1938 2,428,913 Hulst Oct. 14, 1947

Images