US2766377A

US2766377A - Electronic commutator

Info

Publication number: US2766377A
Application number: US321156A
Authority: US
Inventors: Clarence E Frizzell
Original assignee: International Business Machines Corp
Current assignee: International Business Machines Corp
Priority date: 1952-11-18
Filing date: 1952-11-18
Publication date: 1956-10-09
Anticipated expiration: 1973-10-09
Also published as: JPS307754B1; FR1098548A; DE973628C; NL182888B; GB750259A; BE524339A; CH320395A

Description

Oct. 9, 1956 c. E. FRIZZELL EILECTRONIC COMMUTATOR 15 Sheets-Sheet 1 Filed Nov. 18, 1952 xOm KOOM

INVENTOR CLRENCE E. FRIZZELL TTQRNEY 15 Sheets-Sheet 2 Filed Nov. 18, 1952 C. E. FRlZZEl-L ELECTRONIC COMMUTATOR Filed Nov. 18, 1952 15 Sheets-Sheet 5 m m m AAIA AAAA Omm nw OTTLG v INVENTOR CLARENCE E. FRIZZELL ATTORNEY Oct. 9, 1956 c. E. FRIZZELL ELECTRONIC COMMU'IATOR 15 Sheets-Sheet 4 Filed Nov.

Oct. 9, 1956 c. E. FRIZZELL ELECTRONIC coumuwmog 15 Sheets-Sheet 5 Filed Nov. 1 1952 Omm mW INVENTOR CLARENCE E. FRIZZELL W ATTORNEY Oct. 9, 1956 G. E. FRIZZELL ELECTRONIC COMMUTATOR 15 Sheets-Sheet 6 Filed Nov. 18, 1952 G. E. FRIZZELL 2,766377 EILECTRONIC COMMUTATOR 15 Shegts-Sheet. 7

Filed Nov. 18, 1952 TTORNE'Y Oct. 9, 1956 c. E. FRIZZELL 2766377 ELECTRONIC COMMUTA'IOR Filed Nov. 1 1952 15 Sheets-Sheet 8 INVENTOR.

CLARENCE E. FRIZZELL W /W M E) /ITORNEY Oct. 9, 1956 Filed Nov. 18, 1952 c. E. FRIZZELL ELECTRONIC OMMUTATOR 15 Sheets-Sheet 9 ATTORNEI Oct. 9, 1956 G. E. FRIZZELL ELEC TRONIC COMMUTATOR 15 Sheets-Sheet 1G Fled Nov. 18, 1952 III.C I

INVENTOR CLARENCE E. FRIZZELL ATTORNEY Oct. 9, 1956 TPM CHARACTE CYCLE W2 WZ Filed Nov. 18, 1952 C. E. FRIZZELL ELECTRON IC COMMUTATOR 15 Sheets-Sheet 12 TTORNEY IIIII Oct. 9, 1956 C. E. FRIZZELL ELECTRONIC COMMUTATOR Filed Nov. 18 1952 15 Sheets-Sheet 13 FIG.32

ATTORNE R7 WO Oct. 9, 1956 Filed Nov. 18, 1952 c. E. FRIZZELL ELECTRONIC COMMUTATOR 15 Sheets-Sheet 14 ATTORNEY United States Patent ELECTRONIC COMMUTATOR Clarence E. Frizzell, Poughkeepsie, N. Y., assignor to International Business Machines Corporation, New York, N. Y., a corporation of New York Application November 18, 1952, Serial No. 321,156

17 Claims. (Cl. 250-27) This nvention relates to electronic pulse generating circuits and particularly but not exclusively to circuits for generating a sequence of pulses for switching or gating purposes in computing and similar apparatus.

It is sometimes required that periods of time must be subdivided into accurately determined time intervals, each interval being indicated by an electrical pulse. If the number of time intervals in each period of time is equal, the sub-division may be carried out by the use of a closed ring of bistabie trigger circuits. The two conditions of stability of the triggers are designated as on and oi conditions, and usually only one trigger in the ring is in the on condition. A master oscillator is used to generate pulses at a frequency equal to one pulse for each time interval. These pulses are applied to all the trigger circuits and their effect is to convert any trigger which is in the on condition to the ofl condition. The shifting of a trigger from the on to the condition generates a potential variation which is fed to the next succeeding trigger to turn it on. This progression or stepping of the on condition from one trigger to the next succeeding trigger may continue indefinitely since the triggers are connected in a closed ring. The period of time which is subdivided is equal to the pulse repettion rate of the oscillator multiplied by the number of trigger stages in the ring.

In some applications, it is required that certain peri0ds have a different number of equal time intervals than others. For example, electronic computer systems utilizing devices for the storage of information may require less time for reading the information from the device than for writing the information. In such a case, it is desirable in order to save time, to have a lesser number of. time intervals in the read period than in the Write.

period.

Accordingly, it is an object of this nvention to provide time intervals of various lengths and means for dividing those intervals.

Another object is to provide an electronic ring circuit including a series of bistable triggers to provide time division according to a prearranged pattern.

Another object is to provide an electronic ring com prising a chain of trigger circuits and including novel means to vary the eiective number of trigger circuits in the ring.

It is a further object to provide a commutator circuit for distributing signals in a timed relationship and according to a prearranged pattern.

Another object is to provide an electronic ring comprising a chain of triggers and including coupling means for selecting different ones of said triggers for closing the chain.

Another object is to provide an electronic ring com prising a chain of triggers, means to count the number of cycles completed by the ring and means controlled by said counting means to change the effectve number of triggers in the ring after a certain number of cycles have been counted.

Another object is to provide an electronic ring compris ing a chain of triggers, means to count the number of minor cycles completed by the ring and means controllcd by said counting means to change the eiective number of triggers in the ring, a number of said minor cycles comprismg a major cycle.

ther objects of the nvention Will be pointed out in the following description and claims are illustrated in the accompanying drawings which disclose, by way of example, the principle of the nvention and the best mode, which has been contemplated, of applying that principle.

In the drawings:

Figs. 1 and 2 illustrate peaker circuits;

Figs. 3 through 6 illustrate trigger circuits;

Figs. 7 through 14 illustrate cathode follower circuits;

Figs. 15 through 18 illustrate switch circuits;

Figs. 19 through 23 illustrate inverter circuits;

Fig. 24 illustrates an oscillator;

Figs. 25, 26 and 27 taken together, as illustrated in Fig. 34, constitute a complete block diagram of the nvention;

Figs. 28 to 33 inclusive, assembled as illustrated in Fig. 35, constitute a complete timing diagram of signals emitted by certain lines of the devices of Figs. 25, 26 and 27; and

Figs. 36 and 37 are block diagrams illustrating how signals developed by Figs. 25, 26 and 27 may be used to develop other signals.

Wherever shown, unless otherwise indicated in the drawings, the values for the various resistors, inductances and condensers are in thousands of ohms, microhenries, and micro-microfarads, respectively. The term positive and negative potentials, used in the discussion of the circuits, refer to relative values rather than values with respect to ground.

In the novel apparatus of the nvention, a timing ring consisting of bistable triggers is provided. As mentioned above, at least one trigger is in an on condition and the others are in ofl conditions. A master oscillator feeds pulses to all the triggers which step the on condition from one trigger to the next succeeding trigger. As the ring completes a cycle, a signal is emitted to a counter and after a predetermined number of cycles have been counted, the number of efiective trigger stages in the ring is changed by including additional triggers orby excluding some triggers already in the ring. This change in the number of stages is accomplished by coupling the output of a certain designated stage of the chain of triggers, to several switch circuits. These circuits are selectively conditioned or not conditioned depending upon which outputs of which counters or control device is effective. The potental variation developed by the designated stage shifting from an on to an o condition is fed through a thus selectively conditioned switch circuit to restart the ring at a point in the ring varying with the switch selected for conditioning. As the ring completes a certain number of cycles, other switch circuits become conditioned and the ring is thus started at diierent trigger stages. As explained below, the signals emitted by the ring may be mixed with signals emitted by the counters to develop signals delineating different portions of the entire cycle.

Referring to the drwings and more prticularly to Figs.-1 and 2, these circuits illustrate details of peaking circuits, designated P-1 and P-2. The circuit shown in Fig. 1 includes a pentode, which may be of the 5763 type,

ply. The junction point of the 50 millihenry coils and the .047K resistor is g-rounded through an .05 microfarad condanser. An output terminal 6 is connected directly to the plate. The control grid is connected through an 0.1K resistor and a 10 micro-microfarad condenser to an input terminal 3 and through the same 0.1K resistor and a 300K resistor to the junction point of the 50 microhenry coils and. the .047K resistor. The sereen grid is connected throught an 0.47K resistor to the same junction point. The pentode is normally conducting and is cut by the negative swing of an input pulse applied to input terminal 3. As the tube cuts 011, the 50 microhenry coils oscillate to produce an 0.2 microsecond pulse output at output terminal 6. A crystal diode, shunt-ing the 50 microhenry coils serves to clamp the negative excursion of the oscillatiou signal so that only a single pulse output occurs, for each input pulse. The .047K resistor and the .05 microfa=rad condenser form a decoupling circuit so that the peaker, on being pulsed, will not draw a large slug of current from the +150 volt power supply. It is wellknown by these skilled in the art, that the connections and value of the circuit components may be varied. In Fig. 2, for example, a 500 miorohenry coil which is in the plate circuit is damped by a 3.6K resistor. A further diterence over Fig. 1 is that the input condenser has a value of 470 microfarads. The output of Fig. 2 is an 0.5 microsecond pulse, for each input pulse.

Referring now to Figs. 3 through 6, these illustrate electronic triggers designated T1 through T-5, all being of a type commonly known in the art as the Eccles-Jordan trigger. This type comprises two cross-coupled triodes in which the plate P-1 (Fig. 3) is cross-coupled by a 110K resistor in series with a .47K resistor to the grid G2 while the plate P2 is likewise cross-coupled to the grid G1 by a 110K resistor in series with .a .47K resistor. Bach of the 110K resistors is shunted by a micromicrofarad condenser. Both the grids G1 and G2 -are connected, through the .47K resistors, mentioned above, and 330K resistors, to a 250 volt power supply. The cathodes K are grounded while each of the plates P1 and P2 are connected through 3.9K resistors in series with 3.3K resistors to a power supply of +150 volts. Outputs are taken from terminal 7 connected directly to the plate P1 and from terminal 8 which is a tapped output on the plate resistors connected to P2. Inputs may be fed to the grid G1 trom terminal 6 through an isolating diode, or to the grid G2 from terminal 3 to which is connected a diiferentiating circuit (comprised of a 20K grounded resistor and a 33 micro-microfarad condenser connected, as shown) and an isolating diode. The isolating diodes pass only negative signals.

If the left triode is conducting and a negative signal is applied to the terminal 6, the left triode is cut off. As the left triode stops conclucting, the voltage at the plate P1 goes positive, which voltage, through the crosscoupling previously described, drives the grid G2 relatively positive so that the right triode conducts; thus P1 is positive while P2 is negative. This is one state of sta bility of the trigger which may be called the off or normal coudition. In a similar manner, with the right triode conductng, the left triode may be rendered conductive, by the application of a negative pulse of suitable amplitude to the terminal 3, whereupon the right triode stops conducting and the voltage of P2 going positive is applied by the cross-coupling connection to the grid G1 causing the left triode to conduct, P1 becoming negative. This is the second state of stability which may be called the on" condition.

The circuits shown in Figs. 4 through 6 are basically the same as the circuit of Fig. 3. The ditrerences occur merely in the values of the circuit components, and the locatious of the outputs (full or tapped plate resistor), or the use of diflferentiating circuits, on both inputs. One additional exception should be noted. In Fig. 4, there is an additional input terminal 5, which may receive a signal to turn the trigger on.

Figs. 7 through 14 illustrate the details of various cathode follower circuits diagrammatically represented by blocks CF-1 through CF-4 and CF-6 through CF-9. Referring specifically to Fig. 7, block CF-1 specifically includes, a pentode having -its plate connected through an .047K resistor to a +300 volt power supply and through an .05 micro-microfarad condenser to ground. The suppressor and sereen grids are connected directly to the plate as shown. The control grid is connected through an 0.1K resistor and a 24K resistor to ground, through the same 0.1K resistor and a 240K resistor, to a power supply of 250 volts, and through the same 0.1K resistor and a 47 micro-microfarad condenser, to the input terminal 4. The cathode is connected directly to the output terminal 6 and through a lK resistor to ground. The tube is normally cut-ofl, the application of a positive signal input causing a positive output in the cathode output terminal 6.

The CF-2 type, shown in Fig. 8, includes a dual triode, the triode sections operating in parallel. The plates are connected through an .047K resistor to a +300 volt power supply and through an .02 micro-farad decoupling condenser to ground. The grids are connected through separate .47K resistors and a common 240K resistor to a power supply of 250 volts, through the same .47K resistors and a common 24K resistor, to ground, and through the same .47K resistors and a common 47 micro microfarad condenser, to an input terminal 4. T he cathode is connected directly to the output terminal 6 and through a 0.1K resistor to ground. The tube is normally cut-ofl and is driven by the positive pulse output from the peaker P1 of Fig. 1 applied to its input 4. A plus output pulse is available at the terminal 6 for each positive pulse applied to the input terminal 4.

The CF-3 type, shown in Fig. 9, includcs a dual triode with each half acting independently of the other identical half. Referring to the left triode, the plate is connected directly to a +300 volt power supply, the control grid is connected through a resistor to the input terminal 4 and the cathode is grounded through the series connectcd resistors of .2K, 6.8K and 3.9K. The output terminal 6 is tapped between the .2K and the 6.8K resistors. The right half is identical, except that terminal notations of 9 and 7 replace terminal notations of 4 and 6. As the gricl of either half goes positive, the related output goes positive.

Fig. 10 shows a CF-4 type which includes a dual triode with both plates directly connected to a +150 volt power supply. The grids are connected through separate resistors and a common 300K resistor to a power supply of 250 volts and through the same separate resistors each in series with a common 150K resistor, to the terminal 9, the 150K resistor being snunted by a 22 micro-microfarad condenser. Each cathode is connected by a 4.7K resistor in series with a 2.2K resistor to a power supply of volts.

The

output terminals

6 and 7 are connected directly to the cathode and both emit positive outputs when a positive signal is applied to the input terminal 9.

The CF-6 and CF-7 types, illustrated respectively in Figs. 11 and 12, are basically alike, diiering trom one another only in the values of the circuit components'; therefore, explanation is made of Fig. 11 only. The plate of the left triode is connected directly to a volt power supply, its grid is connected through a .47K resistor and a 330K resistor to a power supply of 250 v0lts and through the same .47K resistor and a 150K resistor to the input terminal 4, the 150K resistor being shunted by a 15 micro-microfarad condenser. The cath- 0de is connected directly to an output terminal 6 and through a 6.8K resistor, in series with a 2.2K resistor, to a power supply of 100 volts. A positive input to terminal 4 causes a positive output at terminal 6. The circuit of the right half of the dual triode is identical with that just described, terminal 9 being the input and terminal 7 being the output.

5. TheCF-8 type, shown in Fig. 13, differs:fromthe CF-4 type, described above, only in values of certain components; A positive input to terminal 9results in positive outputs at

terminals

6 and 7.

Fig. 14 illustrates a CF-9 type. Referring to the left half of the dual triode, the plate is connecteddirectly to +150 volt power supply; the grid is connectedthrough a .47K resistor in series with a 360K resistor to the same +150 volt power supply, through thesame .47K resistor, in series with a 20K resistor, to ground and through the same .47K resistor, in series with a 47 micro-microfarad condenser, to an input terminal 4. The cathode is connected directly to an output terminal 6 and through two 2.2K resistors, respectively in series, to a +100v01t power supply. The input is condenser coupled and the circuit operates in class A. Therght half is identical, with terminals labeled 9 and 7 correspond to the terminals labeled 4 and 6 of the left half.

Figs. 15 and 16 show switching circuits, designated respectively S1 and S2, that are identical, except for the value of an input condenser. Each circuitzuses a pentagrid tube having a plate connected to an output terminal 7 and through resistors of 3.9K and 3.3K inseres to a power supply of +150 volts. The cathode andthe number grid are grounded. The number 2 and number 4 grids are connected together and through a resistor to the power supply of +150 volts. The number 3 grid is connected through a resistor, in series with a 510K resistor, to a power supply of 250 volts and through the same resistor and a 240K resistor, shunted by a 33 micromicrofarad condenser, to an input terminal 6. The number 1 grid is.connected through a resistor and a 20K resistor to ground through the sameresistor anda 240K resistor to a power supply of minus 250 volts, and through the same resistor either a 47 micro-microfarad condenser (as indicated in Fig. 15) or a 100 micro-microfarad condenser, as shown to an input terminal 3. The switches S-1 and S-2 operate as follows: The circuit may be coriditioned, by a positive input to terminal 6. A positive signal applied to the terminal 3 is dilerentiated by the input condenser and the associated resistor network so that only a sharp positive pulse reaches the grid. The tube conclucts a short time only, that is for a time interval approximating the duration of the sharp pulse developed by the differentiated signal. Therefore, if the inputto terminal 6 is positive and a positive signal (pulse or steady state voltage) is also applied to terminal 3, a negative pulse is available at the output terminal 7.

The switch S-4 shown in Fig. 17, is identical to switch S-2 except that the plate is not connected throughplate resistors to a power supply. This switch is used to pull triggers to on 01 off conditions and is connected to the output terminal of a trigger in such a manner that the plate circuit of the trigger is also the plate circuit of the switch. Otherwise, switch S4 operates in the same manner as switch S-2.

The switch circuit S-5 (Fig. 18) is also identical with the switch S-2 (Fig. 16) except for the plate circuit. The plate is connected through a 500 microhenry coil in series with a .047K resistor, to a power supply of +150 volts and through the same 500 microhenry coil, in series with .05 microfaradcondenser to ground. The 500 microhenry coil is shunted by two germanum diodes connected in series. This switch operates, so far as inputs. are concerned, in a manner identical to that described above for the switch S-2. However, the plate circuit is that of a peaker and the output at output terminal 7 is asharp negative pulse. The diodes, shunting the 500 microhenry coil, clamp the oscillations of the coil, so that only the iirst negatve pulse appears at the output terminal 7 Fig; 19 llustrates aninverter type I-1.- The plate of the left half qfthe dual triode is connected directly to an outputterminal 7 and through 7.5K and 13K re- Sistors to a power supply of +150 volts. The left grid is connected through a 1K resistor and a 4 70K resistor 6. to a power supply off+250. voltsand through the same 1K resistor and a 200K resistor to the input terminal 5. The right half has an identical plate circuit, with the output taken from the terminal 6. However, the right grid is connected, through a 1K resistor and a 240K resistor, to a power supply of 250 volts and through the same 1K resistor and a K resistor, shunted by a 40 micro-microfarad condenser, to an input terminal 3, the cathodes are grounded. A positive signal input to terminal 5 produces a delayed negative output at termina! 7 and positive signal input to terminal 3 produces a negativeoutput at terminal 6, with relatively no delay since the condenser passes the input signal with relatively no delay.

The inverter type I-3 is shown in Fig. 20. The left plate of the dual triode is connected through resistors er" 3.9K and 3.3K values to a power supply of volts, the output terminal 7 being tapped at a point between the two resistors. The left grid is connected through a .47K resistor and a 20K resistor to ground, through the same .47K resistor and a 240K resistor to a power supply of 250 volts and through the same .47K resistor and a 47 micromicrofarad condenser to the input terminal 5. The cathode is grounded. A positive signal input to. the terminal 5 results in a negative pulse output at terminal 7. The right half of the tube has identicalcircuitry with an input terminal 3 and an output terminal 6.

The inverter I5 shown in Fig. 21 also includes a dual triode. The left plate is connected directly to an output terminal '7 and through a 12K resistor, in parallel with an 18K resistor, to a power supply of +150 volts. The left gridis connected through a .47K resistor in series with a 330K resistor to a power supply of 250 volts and through the same .47K resistor in series with a 150K resistor, shunted by a 15 micrornicrofarad condenser, to an input terminal 5. The cathode is grounded. This is a D. C. inverter and a signal input to terminal 5 is inverted and emitted from terminal 7. The right triode has identical circuitry having an input terminal 3 and an output terminal 6.

The inverter I-6, illustrated in Fig. 22, uses a dual triode with each triode operating in parallel. The plates are connected directly to an output terminal 9 and through a 2.2K resistor to a power supply of +150 volts. The grids are connected through separate 0.1K resistors, and a single 270K resistor, to a power supply of 250 volts, through the same 0.1K resistors and an 18K resister, to ground and through the same 0.1K resistors and a 100 micro-microfarad condenser, to an input terminal 5. The cathodes are grounded. This is an A. C. inverter and a signal input to terminal 5 results in an inverted output pulse at terminal 9.

Fig. 23 illustrates the inverter I-7 which includes a dual triode. The plates have a common connectionwhich leads directly to the output terminal 8, a further connection through an 0.15K resistor to ground and another through three .33K resistors in series to a power supply of +75 volts. The grids are connected through separate 0.1K resistors and a 270K resistor to a power supply of 250 volts, and through the same 0.1K resistor and a 500 micromicrofarad condenser to aninput terminal 6. The cathode is connected, through a selfbiasing circuit consisting of an 0.3K resistor, shunted by an .02 micro-microfarad condenser, to a power supply of 250 volts. A sharp positive pulse input to terminal'6 results in a sharp negative output at terminal 8.

A crystal controlled oscillator F is ill1strated in Fig. 24 Which comprises a pentode whose plate is connected through en 0.5 millihenry coil, shunted by a 10 to 100 micromicrofarad variable condenser, to a +300volt power supply. The sereen grid is connected through a 47K resistor to this power supply of +300 volts and tl'trough an .01 microfarad condenser, to ground. The control grid is connected to ground, through a crystal,

which is shunted by a 1000K resistor. The suppressor and the cathode are connected together and through an 0.25K resistor to ground. The circuit is tuned and controlled by the crystal so that the entire circuit oscillates at a frequency of one megacycle and the output is taken from terminal 9, which is connected to the cathode.

In the particular embodiment of this invention, to be described below, a defined time cycle is divided into periods having various numbers of one-microsecond intervals. The timing diagrams of Figs. 28 through 33 show that a complete cycle designated a character cycle, is divided into groups of G periods, R periods and W periods. These periods may have the following significance when used with an electronic storage device:

During the R times, information may be read from the device, during W times, information may be written in or stored in the device, and during G times, the informaton stored, may be re-generated. The smallest subdivision of the timing chart represents an interval of one microsecond.

It may be seen by inspection of the timing diagram that there is a period of 5 microseeonds (each subdivision, as indicatcd at the top of each Fig. 28 through '34, being one microsecond) from G to G1, from R0 to R1, and from WO to Wi. There are periods of 9 microseconds from G1 to G2, G2 to G3, G3 to G4, and G4 to R0; 4 microseconds from R1 to R2, R2 to R3, R3 to R4, R4 to R5, R5 to R6, R6 to R7 and R7 to WO; 7 microseconds from Wl to W2, W2 to W3, W3 to W4, W4 to W5, W5 to W6, VV'6 to WV7, V7 t0 G0.

Figs. 25, 26 and 27 taken together, show the circuits mainly in block form, for generating various signals, according to the requirements of the timing chart.

At the start of operation, the trggers 122 (Fig. 26) and 61 (Fig. 25) are reset to on conditions, while all other triggers are reset ofi. The reset circuit is not shown but the setting of the triggers may be accomplished by any on of several well known means, for example, by delaying the application of the 250 volt bias supply (sec Figs. 3 through 6) to the left grid of

triggers

122 and 61 and to the right grid, of all the other triggers. The oscillator 30 (Fig. 25) is oscillating at a one megacycle rate and the output therefrom, is fed through the inverter 31, the peaker 32, the cathode follower 33 and the inverter 34 to line 435. Line 435 feeds a series of pulses (sec line 435 of the timing diagram, Fig. 33) to the input terminal 6 of

triggers

61, 62, 63, 64, 65, 67, 68, 69, 70, 71, 72, 73 and 74. These triggers are connected in cascade to form a main chain, each stage of the chain having two conditions of stability. Under the assumed conditions, the trigger 61 is in an on condition while all other stages of the chain are in 0 condi tions. The first negative excursion of the signal (trailing edge) on line 435 shifts the trigger 61 ot but has no effect on the other triggers of the chain, the latter triggers already being ot'f. As the trigger 61 shifts off, a negative output is developed at terminal 8 and is fed to terminal 3 of trigger 62, which shifts 62 on.

As trigger 62 shifts on, a positive signal is emitted from its terminal 8 to the input terminal 3, of trigger 63, but it has no etect, since the diode in the input circuit (sec Fig. 3, for example), blocks passage of positive signals. The next negative going signal on line 435, occurring one microsecond after the first, shifts trigger 62 ofl, in turn, shifts trigger 63 on. Thus, successive signals on line 435 cause the operation of the chain of triggers to progress, by shifting the on condition from one trigger to the next, in each of the successive triggers of the chain. For reasons that will be better understood later, trigger 61 is called the 9 stage of the chain, trigger 62 is the 8 stage, trigger 63 the 7 stage, trigger 64 the 6 stage, trigger 65 the 5 stage, trigger 67 the 4 stage, trigger 68 the 3" stage, trigger 69 the 2 stage, trigger 70 the 1 stage, trigger 71 the 0" or index stage, trigger 72 the +1 stage (signifying one micro- 8 second after the 0 or index time), trigger 73 the +2 stage, and trigger 74 the +3" stage.

The outputs of the triggers are made available at the fcllowing points: Trigger 61 through cathode follower to line 422, trigger 62 through cathode follower 51 to line 423, trigger 63 through the cathode follower 52 to line 424, trigger 64 through the cathode follower 41 to line 425, trigger 65 through cathode follower 53 to line 426, trigger 67 through cathode follower 54 to line 427, trigger 68 through cathode follower to line 428, trigger 69 through cathode follower 42 to line 429, trigger 70 through cathode follower 56 to line 430, trigger 71 through

cathode followers

57 and 44 to line 431, trigger 72 through cathode follower 58 to line 432, trigger 73 through cathode follower 59 to line 433, and trigger 74 through cathode follower to line 434. These outputs on the various lines are shown on the timing diagram with corresponding numbers and will be referred to below as the outputs of the triggers.

After the start of operation as described above, and as the on condition is stepped from the trigger to trigger 71, the cycle has reached G time (sec timing diagram). The leading edge of the negative pulse output from terminal 7 of trigger 71 is fed through the cathode follower 57, the peaker 36, the cathode follower 37 and line 37A to terminals 3 of

switches

80, 82, 83, 85, 86, 88, 89, 90, 91, 92, 93 and 95. At the G0 time, under the reset condition previously mcntioned, the trigger 122 (Fig. 26) is in an on condition. The output of trigger 122 terminal 7 is available via cathode follower 114 and line 401. Its other output terminal 8 is thcn positive and is fed through the cathode follower 118, the inverters and 106 to terminals 6 of the

switches

86 and 88. The

switches

86 and 88 are conditioned by the latter input and the pulse occurring on common input line 37A at the G0 time, is passed tl1rough these conditioned switches. T he output from terminal 7 of switch 88 is negative and is coupled through 33 micro-microfarad condenser 87 to terminal 5 of the trigger 65, the latter being thus shifted on.

At the same time, the output of switch 86, terminal 7 is fed via line 86A to terminal 6 of trigger 122, shifting the latter ofi. The output of trigger 122 terminal 7 goes positive and is fed via the inverter 117 to terminal 3 of the trigger 120, shifting that trigger on. The trig ger 120 terminal 8 goes positive and via line 120A and

inverters

101 and 102, conditions the

switches

80 and 82. The output of trigger 120 terminal 7 goes negative and via line 120B turns on the trigger 162.

Therefore, during the first microsecond, after the G0 index time, triggers 120, 162, 65 and 71 are on" while all other triggers are 011. The

switches

80 and 82 remain conditioned, as long as the trigger 120 is on."

The next impulse on line 435 occurs during this first microsecond after G0 time. The negative excursion (the trailing edge) steps the on condition from the 5" stage trigger 65 to the 4 stage trigger 67 and from the 0" stage trigger 71 to the +1 stage trigger 72. However, the shifting of the conditions of the triggers are not completed, due to circuit delays until the end of the microsecond divisions shown on the timing diagram. Therefore, th outputs of trigger 65 and 71 (

lines

426 and 431 of the timing diagram) are one microsecond negative signals occurring immediately after G0 time.

The stepping action continues and for the second mi crosecond after G0 time, triggers 67 and 72 are on, then triggers 68 and 73, etc., until trigger 71 is again turned on. The operation of the device has now progressed to the G1 time.

The negative output from trigger 71 terminal 7 again passes through the cathode follower 57, the peaker 36 and the cathode follower 37 to line 37A. As described above, the trigger 120, via line 120A and

inverters

101 and 102 is conditioning the

switches

80 and 82. The signal, now present on line 37A, passes through the switch 82 and the 33 micro-microfarad condenser 81 to shift the trigger 61 on and also through the switch 80 and the cathode follower 96to line 96A.

Triggers

142, 150, 156 and 162 are connected as a chain similar to the main chain described above, and thesignal now present on line 96A serves to step the on condition trom one trigger to the nextsucceeding trigger. As explained anove, trigger 162 was turned on at G0. time, and therefore, trigger 162 is now at G1 time, turned off and trigger 156 is turned on. The output of trigger 162 terminal 7, negative from 30 to G1 time, but now positive is fed through the cathode follower 163 to line402. A plot of this output is also shown on the timing diagram.

The main chain continues to be stepped from the 9 stage trigger 61 on and requires nine microseeonds before another negative output is available from the trigger 71. .Inspection of the timing diagram shows that this output occurs at G2 time and after passing through circuitry, previously described, to line 37A, it reaches switches 80 and 82. Since the trigger 120 is still on, the

switches

80 and 82 are cOnditioned.. The output from switch 82 again shifts the trigger 61 on while the output from switch 80 via the cathode follower 96 and line 96A steps.an on condition from trigger 156to trigger 150. Trigger 156 has been on trom G1 to G2 time and its output is available via cathode follower 157 and line 403.

The man chain continues to be stepped. At G3 and G4 times, through the circuits and under conditions just explainedthe 9 stage trigger is set to an on condition and,the sub-chain receives stopping signals. At G3 time, the On condition is stepped in the sub-chain from trigger 150 to trigger 142. However, at G4 time, when trigger 142 is shifted o, a signal is emitted from its terminal 8 via line 142A to turn 08' the trigger 120 and to turn on the trigger 123. The trigger 120 being o, removes conditioning from the

switches

80 and 82. The trigger 123, now on emits a positive output from its terminal 8 via cathode follower 115 and

inverters

107 and 108 to condition the

switches

89 and 90. The output of trigger 123 terminal 7 is available via cathode followers 130 and 131, and line 406.

Niue microseconds later at R time, as the main chain completes another cycle, the signal trom trigger 71 reaching line 37A passes through the now conditioned

switches

89 and 90. The output trom switch 90 via the 33 micromicrofarad condenser 87 resets the stage trigger 65 to an on condition. The output of switch 89 output 7, via line 89A turns off trigger 123. The output of trigger 123 terminal 8 via the cathode follower 115 and line 115A turns on the trigger 125 and via the same cathode follower 115 and

inverters

107 and 108 removes the conditioning from the

switches

89 and 90.

The trigger 125 now on emits a positive signal from terminal 8 via line 125A and inverters 111 and 112 to condition the

switches

93 and 95 and a negative signal from terminal 7 via line 125B turns on trigger 178. Trigger 178 is the first stage of a second sub-chain comprised of

triggers

178, 174, 170, 166, 160, 154 and 147. The outputs of this second sub-chain are available as follows: trigger 178 via cathode follower 179 and line 407; trigger 174 via cathode follower 175 and line 408; trigger 170 via cathode follower 171 and line 409; trigger 166 via cathode follower 169 and line 410, trigger 160 via cathode follower 161 and line 411; trigger 154 via cathode follower 155 and line 412; and trigger 147 via cathode follower 149 and line 413.

Five microseconds later, at R1 time, the output of trigger 71 reaching line 37A passes through the conditioned switches 93 and 95. The output of the switch 95 via the 33 micro-microfarad condenser 94 turns on the 4 stage trigger 67, while the output of switch 93 via cathode follower 98 and line 98A steps the second 10 sub-chain, to turn ot trigger 178 and turn on trigger 174.

Ths operation, of resetting the 4 stage trigger 67 to an on condition and stepping the on condition along the second sub-chain, is repeated at R2, R3, R4, R5and R6 and R7 times. However, at R7 time, as the trigger 67 is set to an on condition, the trigger 147 is turned otf and the output of its terminal 8 via line 147A, turns off trigger 125 and turns on trigger 124. Trigger 125, going 0, removes the conditioning from the

switches

93 and 95, while the output of trigger 124, terminal 8, via cathode follower 116,

inverters

109 and 110 and line 110A, conditions theswitches 91 and 92. The output of trigger 124 terminals 7 is available via cathode follower 133 and line 414.

Four microseconds later, at WO time, the trigger 71 is turned on and the resulting signal on line 37Apasses through switches 91 and 92. The output of switch 92 viathe 33 micro-microfarad condenser 87, sets the 5 stage trigger 65 to an on condition, while the output of the switch 91, via line 91A, turns off the trigger 124. The trigger 124 going otf, removes the. conditioning signal from the switches 91 and 92 and its output at terminal 8 via line 124A turns on the trigger 121. The output of trigger 121 terminal 8, via

inverters

103 and 104, now conditions the

switches

83 and 85, while the output of terminal 7 via line 121A turns on the trigger 176 which is the first stage of a third sub-chain, including

triggers

176, 172, 168, 164, 158, 152 and 144. The outputs of these triggers are available as follows: trigger 176 via cathode follower 177 to line 415, trigger 172 via cathode follower 173 to line 416, trigger 168 via cathode follower 169 to line 417, trigger 164 via.cathode follower 165 to line 418, trigger 158 via cathode follower 159 to line 419, trigger 152 via cathode follower 153 to. line 420, and trigger l44via cathode follower 146 to line 421.

The next index time Wl, occurs five microseconds after W0 time, and the signal on line 37A passes through the

switches

83 and 85. The output of switch via the 33 micro-microfarad condenser 84 sets the 7 stage trigger 63 to an on condition, while switch 83 emits a signal, via cathode follower 97 and line 97A to step this third sub-chain, specifically, at Wl time, to turn ot trigger 176 and turn on trigger 172.

Ths operation is repeated every seven microseconds, at index times of W2, W3, W4, W5 and W6. However, at W7 time, the trigger 144 goes off and its output trom terminal 8 via the cathode follower 141 and line 141A turns oi trigger 121 and turns on trigger 122. Trigger 121, going 0 removes conditioning trom switches 83 and 85 while trigger 122, terminal 8, now conditions, via the cathode follower 118 and inverters and 106, the

switches

86 and 88. Seven microseconds later the index point of G0 is again reached and the entire operation may be repeated. The conditions now existing are the same as these existing and described at the previous G0 time.

It may be desirable during certain types of operations, to remain in regeneration or the G portion of the cycle and during such times, a positive signal applied to line 126A (Fig. 26) causes the system to operate as follows: As the cycle progresses to the R0 index time, a positive signalonthis line 126A passes through the inverter 126 to terminal 8 of the trigger 125. Ths signal to terminal 8, clamps the. trigger in an ot condition, and the signal at R0 time, normally arriving at pin 3, trom trigger 123, via the cathode follower 115, has no etect. The positive signal on line 126A also condtions the switch 132. Therefore, as the trigger 123 goes otf (at R0 time) a signal is emitted trom its pin 7 via the cathode follower andthe now conditioned switch 132, to turn on the trigger 120. As trigger 120 goes on, the same conditions prevail that exist a fraction of a microsecond, after G0 time. Therefore, the operation of the cycle is switched,