BE494586A - - Google Patents

Description

       

    <Desc / Clms Page number 1>
 



  COMPOSER MACHINE
The present invention relates to a typed typing machine using photographic means for printing.



   The invention will be described element by element, in order to facilitate the description, in relation to the appended drawings in which: FIG. 1 is an overall diagram; FIG. 2 represents the recorder; Figures 3 and 4 show the film carrier; Figures 5 to 9 show the variable escapement; Figure 10 shows the decoder; FIG. 11 is a general diagram of the electrical circuits; FIGS. 12 and 13 represent circuits with 2 stable positions (flip-flop); Figures 14 and 15 show counting circuits; FIGS. 16 to 21 represent in detail the circuits corresponding to FIG. 11; FIG. 18 more particularly represents the correction device s; Figure 22 is an operating diagram of the cams;

   Figures 23 to 25 show a mechanical counter; FIGS. 26 to 29 represent a variant of the decoder; Figures 30, 31 and 32 show variations of the justifier; Figure 33 is an assembly diagram of Figures 16 to 21;

  <Desc / Clms Page number 2>

 FIGS. 34 to 37 represent a variant of the photographic device.



   Referring to Figure 1, the machine described by way of example comprises a CI keyboard associated with a typewriter giving at the same time an easily readable copy on which the operator can read himself and make corrections, an ENR recorder, two counters CP1 and CP2, a JUS justifier, a CHA film carrier, a DEC decoder and a PH photographic unit.



   The CL keyboard is fitted with CM selector bars (also called code or permutation) of the kind used in teleprinter. When a key on the keyboard is lowered, one or more of these selector bars are actuated according to a code similar to that used in Baudot telegraphy. In the example described, a code with 7 elements is used and consequently 7 selector bars which can give 2 or 128 different combinations, which is more than sufficient.



   Each character is characterized by a combination of this code and the permutation bars transmit the combinations to a recorder, preferably mechanical, which records them as you type.



   The writing part of the recorder is driven by the carriage of the typewriter so that in the event of a correction one can operate as with a normal typewriter since the fact of returning the erroneous character to position keystroke brings the writer into place and an erasing device which can be actuated to reset the recorder to zero. The operator then types the correct character instead of the deleted character,
The recorder is simultaneously used for typing and printing control by means of an independent reading device.



   To justify the lines, that is to say to increase the distance between the words so as to align the right margin, it is necessary to measure the length of the line and to count the number of spaces within. - tifiable. To this end, the width of the various characters is expressed as a function of an arbitrary unit of length and a counter CP1 adds up these widths during typing. A second counter CP2 counts the number of justifiable spaces. These two counters are associated with the keyboard during typing and for a short time for the end of the line with the JUS justifier which determines the justification increment to be added to each space during printing.



    According to features of the invention, the width of the characters is determined by a certain number of the elements of the registration code, for example the first 4 in binary numbering. Four elements make it possible to characterize 16 different widths ranging from 0 to 15, which is sufficient in practice.



    When the operator has finished typing a line, the reader of the recorder is associated with the photographic assembly and with a translator of width TL. The photographic assembly comprises a disc or drum carrying the matrix characters which rotates continuously. A DEC decoder controlled by the recorder causes an excessively brief flash of light to emerge at the right moment which photographs the selected character on the film carried by the CHA carriage. The film carriage is then advanced a distance equal to the width of the character under the control of the width translator TL. cator JUS. The length of the justifiable spaces is chosen by the justifi- Let J be the desired justification (in the chosen unit of length).



  When the composition is complete, the addition of the individual widths of the characters (plus the minimum width of justifiable spaces) gives a lower value from D to J, the spaces must be increased by distributing

  <Desc / Clms Page number 3>

 D among them, on the condition, however, that each space can only be increased by a whole number of units (this condition is dictated by the operating mode of the film carrier which can only advance by whole number of units for mechanical reasons). If N is the number of spaces, the D / N ratio will generally not be an integer. The distribution of the D units between the N spaces is done by adding a certain number Q of units to a first group of spaces and Q + 1 units to a second group of spaces.

   The justification mechanism comprises the counter in which the length of the unjustified line has been recorded and to which the number of spaces is added a number of times in succession. The counter thus records the following totals in succession:
J - D (line not justified)
J - D + N (first addition)
J - D + 2N (second addition) etc ... These additions are repeated until the total reaches or exceeds the value J assigned to the justified row. At this time, a mechanism is actuated to effect the distribution of the units between the spaces.



   A feature of the invention resides in the use of the binary numbering system. As well known, this numbering system has only two digits C and 1, so that successive integers are represented by 0, 1, 10, 11, 100, etc ... (corresponding to 0 respectively , 1, 2, 3, 4, etc ... in the decimal system). The binary system is preferable for a number of reasons: first, because it only requires 2-position relays or switches, such as pulled in or idle, or an open or closed circuit, representing only the conditions "yes or no" or "0 or 1".

   Another advantage is that subtractions can be done more easily than in other systems. Subtractions are necessary for corrections and can be done in binary system by a relatively simple inversion process.



   Another characteristic of the invention resides in the photographic composition system making it possible to place on the film with very great precision the image of a character animated by a rapid movement, this precision being independent of the variations in the position of the characters. - teres with respect to each other on the die-carrying disc.



   Another characteristic of the invention resides in an automatic correction system which makes it possible to replace in a recorded line a character of a certain width by a character of another width, without affecting the justification, and at the same time to mechanically delete the character to be replaced or deleted from the recorder.



   Other characteristics and subjects of the invention will appear in the text and the accompanying figures.
The recorder shown in Figures 2 and 18 is a seven element recorder. In order to register a line, it is therefore necessary to have as many rows of 7 pins as there can be characters in a line.



  90 for example. Each of these pins can take two positions, an advanced position or rest position and a depressed or working position.



  The recorder can be divided into three parts, the EB spindle field, the EM hammer group and the EX scout carriage. The pin field consists of two groups of 7 rows. By moving it in the vertical direction by the distance which separates two consecutive rows of pins, one of said groups of 7 rows of pins is associated with the hammers and the other with the explorer. In this way, the operator can record a line while the previously typed line is photographed.



   The field of pins EB comprises two flanges 100E and 101E connected by spacers 112E. These flanges are provided with openings in which the pins B1 ..... B7 can slide. These spindles can take two positions, a rest position B'1 and a work position B1. We

  <Desc / Clms Page number 4>

 has shown in the drawing a single row of pins for clarity of the drawing, but there are in fact as many rows of 14 pins as there can be characters or signs in a row. The distance separating two consecutive pins in the horizontal direction is equal to the pitch of the typewriter.
 EMI4.1
 



  Panai the 14 pins shown in the drawing, seven- cooperate with the typewriter: B'l, B'2, B'3, B'4? B'5, B'6, B'et, sept (Bl..a..eB7) with the explorer cart and the camera. Thanks to this arrangement, it is possible to type a line while the previous one is photographed, and that with a minimum of elements.
 EMI4.2
 



  When a typewriter key is depressed a number of pins are depressed depending on the code combination assigned to that character. In the case of the figure, pins B'2, B'4, B'6 and B'7 are in their working position, that is to say the combination corresponding to the character is 1-4 -6-7 which is the asterisk in the example described (see figure 10). The pins are driven by hammers Ml ... M7 sliding
 EMI4.3
 between two flanges 110E and IIIE. These are normally urged towards their rest position by springs (not shown). The flange 110E which carries this set of hammers is fixed on the carriage of the typewriter so that it positively follows all the movements of said carriage.



  As the carriage of the typewriter moves, it thus presents the hammers successively in front of each of the vertical rows of spindles.
 EMI4.4
 The hammers are operated by RPl.o. tP'ia These palettes work under the control of the CM selection bars, associated with the typewriter keyboard. When a keypad key is depressed, the bars corresponding to the code of the desired character are actuated and depress the pins.
 EMI4.5
 ches. In Figure 2, the control paddles RP2, RPI, RP6 and RP7 are shown in their operating position. The control of these pallets can be either electrical or mechanical.

   In the case of an electric control, the actuated selection bars close the electro-
 EMI4.6
 magnets RPE1, RPEz u 00 U RPE which rotate the pallets RB1 aa.aRB7.



  When the operator has finished typing the line, he moves the field
 EMI4.7
 of pins in the vertical plane along guides -U3E. This field of pins can take two positions, a low position and a high position. In the case of figure 2, it is in the high position, that is to say that the
 EMI4.8
 odd pin lines (counting from the top BI .. 0 0 "B7), on which the previous line was recorded, are associated with the explorer carriage, and even pin lines (B'1 .... oB 'ï) are associated with the hammers. In figure 2, pins B1, B4 and B5 of the odd rows were inserted during the typing of the previous line, this combination corresponds to the character (see figure 10) in the exemplary embodiment described.

   Organs intended to detect the registered combination or feelers
 EMI4.9
 are represented in Plu .P8 ,. They can rotate around the axes 103E. These axes are carried by arms 115E fixed to a plate 104E. This plate 104E slides in brackets 105E and is pushed towards the arm by a spring (not shown). For the search for the combination, the plate 104E is lifted by means of the pivoting pallet 109E controlled by the electro PLP (figure 18).

   The feelers are thus pushed upwards and those which meet the pins in the depressed position on their travel tilt clockwise and close the contacts which are assigned to them.
 EMI4.10
 companies Cl ---- C8- It is these contacts which control the printing of the corresponding character as will be explained in detail later. The explorer wagon read is provided with rollers 107E running on rails lOBE and an escape mechanism of the typewriter wagon type, controlled by an electrOEXP. After having been "explored" in this way, the pins are pushed back into their rest position by a vertical oscillating bar (not shown).

   The pins are held in their position by springs.
 EMI4.11
 friction of the springs Although the explorer trolley EX and the set of EM hammers have been shown opposite each other to simplify the figure, they move absolutely independently of each other. EM is controlled by the keyboard while EX is controlled by the photographic assembly.

  <Desc / Clms Page number 5>

 



  Truck and its controls - Figure 3 shows a partial section -
 EMI5.1
 the of the drive mechanism of the sensitive film carrier, and FIG. 4 is an elevation of this carriage itself.



   The sensitive film passes from a debtor magazine to a receiving magazine (not shown), supported by a stationary member or a movable carriage but preferably not rigidly connected to the actual carriage shown (figures 3 and 4). . The independence between these two organs is ensured by a sufficiently loose loop formed by the film. When it leaves the debtor magazine, the film loops, passes over debtor 1-F, then under presser 3-F, then under debtor 2-F and returns to the receiving magazine.
The movements of the carriage are determined by an irreversible screw.
 EMI5.2
 relatively fine pitch sible 17-F, the rotation of which is controlled by the variable escapement described later.

   The wheel 37-F (figures 5 to 9) of the variable exhaust turns, after each printing of a character, by an angle determined by the width of this character., This wheel meshes with a disengageable pinion 19-F (figure 3), which itself engages with a 20-F pinion pinion on the 29-F shaft which carries the 17-F screw. Tapered bearings 21-F and 22-F keep the shaft 29-F in translation, so that its rotation communicates to the carriage, thanks to a nut E, a translational movement in the direction of arrow F.



   To ensure a constant gap between the characters, the movements of the carriage must be done with great precision, of the order of 0.01 mm. In order to obtain this precision, it is necessary to make up the clearances, which is achieved by means of two springs which are also used to bring the carriage back after printing a complete line. One of these springs is the 23-F barrel spring (figure 4)
 EMI5.3
 which acts on the carriage by 1-lintermediate the rope 24-F, and which tends to rane- ner this carriage in the direction opposite to the normal direction of movement. The other spring acts on the screw 17-F by means of a 25-F wire and tends to turn this screw in the opposite direction to the normal direction of rotation.



   Each time a character has been photographed, the cart will move a varying distance in a relatively short time, which highlights
 EMI5.4
 play of important energy forces. To ensure the necessary precision, it is essential to absorb most of the shocks caused by stopping the truck.

   A resilient device is provided for this purpose, consisting of a ring 26-F which can slide in the frame of the machine, and accommodating the tapered roller bearing 22-F, this ring being held in place by a spring. powerful 27-F, the tension of which can be adjusted by acting on a 28-F part screwed into the frame.,
Thus, when the variable escapement finishes its movement, the
 EMI5.5
 wheel 37-F (figures 5 to 9) stops abruptly and, due to the inertia, the carriage tends to continue its movement, which applies a translational force on the screw 17-F and on the shaft 29-Fa This force manifests itself on the thrust bearing 22-F by an impact, transferred to this bearing by a shoulder.
 EMI5.6
 ment 30-F of shaft 29-F,

   then transmitted to the ring 2C-F which moves very slightly compressing the spring 27-Fe mant DEB When a line has been completely photographed, the electromagnet DEB is excited. Despite the antagonistic action of the spring 31-F this electro
 EMI5.7
 attracts the plunger 32-F integral with the pinion 19-king, and the latter slides on a rod 33-F fixed to the frame of the machine; it emerges from
 EMI5.8
 wheel 37-F while remaining in mesh with pinion 20-F. As soon as the pinion 19-F is released from the wheel 37-F, the spring controlling the rope 25-F turns the screw 17-F in the opposite direction of its normal direction of rotation, which brings the carriage back into position. back to print the next line.
 EMI5.9
 



  T Ue 37-P In normal operation, the pinion 19 F is kept in engagement with the wheel 37-F by a rod 34-F which slides freely in a fixed core 35-F and which is constantly pushed by an adjustable spring 31- F.



   The MP disc is used for the precise positioning of the cha-

  <Desc / Clms Page number 6>

 riot, as will be shown below.



   An improved variable escape mechanism for controlling the advance of the film carrier will now be described with reference to Figures 5 to 9.



   This variable exhaust comprises a toothed wheel 37-F meshing with the pinion 19-F which controls the rotation of the screw 17-F to advance the carriage. This wheel advances by a number of teeth which depends on the width of the character which has just been reproduced, under the control of a star part 38-F whose movement is limited by stops 41-F named - bre equal to the various advances desired. Star 38-F is provided, in addition, with a toothed sector 39-F having the same axis as wheel 37-F and star 38-F, and which can mesh with a pinion 40- F.



   The width translator of the TL machine first places, at the appropriate location, a stop 41-F, for example by means of an electromagnet (not shown). Then the main electromagnet 42-F is energized and attracts a control rod 43-F. This rod is provided with two stops 44-F and 45-F and passes through an opening made in the tail 46-F of the lever of a claw 47-F;

     this lever can pivot around a 48-F axis. At the start of the displacement of the rod 43-F, a spring 49-F pushes the lever of the claw 47-F, keeping it pressed against the stop 44-F, and the teeth of the claw engage with those of the wheel 37-F and thus immobilize this wheel.



   The rod 43-F continues its course, and its stop 50-F pushes the tail of the support 51-F of the pinion 40-F (FIG. 6); this support swings around its axis 52-F and releases the pinion 40-F from the toothed wheel 37-F. The pinion 40-F then engages the teeth of a stop 53-F which holds it in position.



   At the same time, the star 38-F whose toothed sector 39-F was engaged with the pinion 40-F is released and moves counterclockwise under the action of a spring ( not shown), until one of its branches meets the stop 41-F (figure 7). At this time, the rod 43-F has finished its travel and the stopper 45-F maintains the claw 47-F, which itself retains the toothed wheel 37-F. The pawl 54-F, which is normally pushed into the teeth of the 37-F wheel by a 55-F spring, and which has risen during the rotation of the star, falls back into a tooth of the 37-F wheel and prevents any bouncing of the star.



   A 56-F damper is normally repelled by the star's 57-F tail. When the latter rotates, this damper rotates around the axis 58-F under the action of a spring 59-F and abuts against a stop piece 60-F.



   The mechanism remains in this position until the electromagnet 42-F releases. At this moment, the rod 43-F comes back under the action of the spring 61-F (figure 8). At the start of this return, the part 62-F of the stop 50-F meets the tail 63-F of the part 51-F and this part tilts; the pinion 40-F which it carries meshes both with the toothed wheel 37-F and with the toothed sector 39-F of the star. The part 62-F slides under the shank 63-F and thus keeps the pinion 40-F engaged.



  Continuing its return movement, the rod 43-F releases the claw 47-F, the teeth of which protrude from those of the wheel 37-F (figure 9). At this moment, this wheel urged by the main spring of the carriage (not shown) turns clockwise and drives the star 38-F until the tail 57-F of this star abuts against stop piece 56-F. This is because gear 37-F and star 38-F move at the same time, since they are both in mesh with pinion 40-F. In passing, the tail 57-F gives up part of its kinetic energy to the shock absorber 56-F which, under the shock, takes the position 56'-F.

   To reduce the violence of the impact, a fairly powerful spring is also provided in the stop piece 64-F. The object of this decoder is to generate, when the selected character passes over the optical axis, a re- pulse.

  <Desc / Clms Page number 7>

 
 EMI7.1
 laterally wide which acts on a "window" el.ectroic tube and allows the photoelectric position pulse corresponding to the selected character to pass.



  This decoder is made up of a cylinder comprising 7 rows of isolating and conducting parts on which 7 Ba-l .... Ba-7 brushes rub.



   The combinations of insulating parts and of conductive parts along a generator of the cylinder correspond to the code combinations of the various characters to be reproduced, as shown in FIG. 10, where the diagram shows a developed decoder cylinder. On the other hand, the
 EMI7.2
 BA brushes are each connected to a relay contact Tl to T7 (figure 17) which, under the control of the recorder, is opened or closed according to the code of the character considered, as we have seen previously, this code being the same as that of the combinations of insulating parts and conductive parts of the decoder cylinder.



   To facilitate understanding of the operation of the decoder, this operation will be described on a particular example, considering, as an indication, that we want to reproduce the letter t which, in the arrangement shown in Figure 10 where the conductive parts are hatched, occupies the generator 27 of the cylinder. The figure shows that the combination
 EMI7.3
 corresponding to this letter is 1235, that is to say that when the generator 27 of the cylinder which moves in the direction S passes in front of the brushes BA, the brushes Boa-1, BA-2, BA-3 and BA -5 only rub on conducting parts of the cylinder.



  If it is assumed that the contacts ti, t2, t and t are open, by actuating the corresponding relays, examination of figure 10 shows that, while the cylinder moves in front of the brushes from the zero position to the position 27, at least one of the brushes BA is in contact with a conductive part at a given instant. Consequently, the circuit going from the mass M of the decoder to the common point H (FIG. 10) is always closed. On the contrary, this circuit is open when the brushes pass to position 27. In fact, at this time, the combination of the open contacts is precisely the same as that of the conductive parts of the decoder.

   The circuit therefore remains open for the duration of the passage of the brushes on row 27, and this cut-off is used, as will be seen later, to allow the photoelectric pulse generated at the precise moment of the passage of the letter t carried by the disk D in front of the photocell; this impulse can thus
 EMI7.4
 if puncture the flash tube and produce 3.9iz.presszon photogTaph1- que on the film, of the desired character, ëpu-4 It will be noted that for other subsequent rows of the decoder, there are combinations which again open the aforementioned circuit; thus, in the example chosen, it can be seen that this circuit will still be open when passing under the brushes BA of the generators 39, 50, 53, 58, 63, 65,
 EMI7.5
 fls 753 76, 79, 81, and 82.

   To prevent the characters corresponding to these generators from also being photographed when the letter t is desired, according to one characteristic of the invention, a device is provided which allows the flash tube to operate only once per complete revolution of the decoder. , so that the pulses which occur after Correct trigger pulse have no effect. u. L
 EMI7.6
 ls be. Line counter - The keys of the typewriter activate the selection bars which close their contacts according to the code assigned to the various characters, We will consider here the first 4 contacts only
 EMI7.7
 which characterize the width of the characters. These A first contacts are connected to the 4 sRtEe.es of the binary counter.

   It is clear that these 4 inputs must be able to actuate simultaneously since the contacts of the selector bars are simultaneous. The result of the addition of these numbers in the meter is recorded so-called accumulator or binary relays, or combinations of relays. These relays have two stable positions and they move from one position to another when current is applied to them and then in. - interrupted, and remain in this position. For this reason they are also called flip-flope. FIG. 14 represents an exemplary embodiment using a step-by-step switch mechanism, COmm.tU1é.ment used in telephony. This

  <Desc / Clms Page number 8>

 switch includes a pawl 44 and a ratchet wheel 46.

   When the electro 48 is energized, the armature 50 urges the pawl back against the action of a spring 54 and when the current is interrupted the ratchet wheel advances the contact disk 54 of. a step. A brush 56 makes permanent contact with the disk near its center and another brush 58 is in contact with a metal part, for the positions where the circuit is closed, and with an insulating part for that where the circuit is open. It follows that the circuit is closed every two times that the electro 48 is energized. An auxiliary winding 60 may be provided for a purpose which will be discussed later.



   Figure 15 shows a flip-flop circuit which gives the same results but uses normal telephone relays instead. step-by-step more expensive. In figure 15, when earth is applied through key S, current, passes through relay 62 and one winding of relay 64 in series, but a single winding of 64 is too weak to actuate the relay. relay 64. (The term actuate is used here to denote the operation of the relay armature). As a result, the relay 62 operates alone. When the key is opened, the second winding of 64, which during the pulse was grounded on both sides, is then energized.

   These 2 windings of 64 and the winding of 62 remain energized and the two armatures are attracted. When the key is closed, a second time, the 2 ends of the 62 winding are connected to earth and it releases, but 64 remains actuated by its 1/2 winding (this relay is of the well known type of double relay winding in which a single winding is unable to actuate the relay but can keep it actuated once it has operated).



   When the S key is opened, 64 releases and the cycle can repeat. The contacts 66 controlled by the relay 64 can be used in an external circuit. These contacts close every two operations of the key S and therefore function in the same way as the brushes 56 - 58 of figure 14.



   If we now consider a binary relay of a given stage, it is clear that it must operate when a pulse is applied to it and that it must transfer this pulse to the upper stage if it is found. already in its actuated position or tilt! position, and that it must not transfer this restraining pulse if it is in its "0" position. However, if a stage receives both a carry pulse from the lower stage and a direct pulse, it should not operate but must transmit a carry to the next higher stage, regardless of which position it is in. initially found.



    FIG. 12 represents a binary counter fulfilling these conditions. It includes as many flip-flop circuits as there are floors (5). These circuits are represented by X1, X2, X3, X X5 and can be of the type shown in the dotted box of figures 14 or 15. Each flipflop circuit is provided with contacts which are closed when the flip-flop circuit is in position "1 "and open when in position" 0 ". The contacts are shown at 66. Such simple closing contacts on the flipflop circuit would be sufficient if the inputs to the various stages were not simultaneous but occurred in succession.



  When a flip-flop circuit receives a hold potential from the previous stage, and a direct operating potential from its own stage, it is impractical to operate that flip-flop circuit twice. A faster and more convenient method is to prevent the flip-flop circuit of one stage in which there is a double input from operating, and to send it to the upper stage a carry potential.



  The circuit of FIG. 12 fulfills these conditions. All stages are identical, except the first which, of course, cannot receive any holding potential. Each stage is fitted with an R2, R, etc. restraint relay.



  These relays have two windings in opposition and operate when any one of their windings is energized and remain at rest when both windings are energized at the same time. The entrance to the first floor is

  <Desc / Clms Page number 9>

 made by a wire 67 having a contact (represented by the key K1), the said con-
 EMI9.1
 driver is directly related to X 1 and also to è. one of the contacts 66. A restraining conductor 68 runs from contacts 66 of X1 to a winding of j. Input wire 68 goes from contacts 66 of X1 to a winding of IL.

   The second stage input wire 69 goes through JS to the second winding of IL and also through a rectifier 70 and wire 71 to contact 66 (but not to the flip-flop circuit] L itself) A second rectifier 72 is connected between wires 68 and r rJl The relay IL has its working contacts 73 connected by a wire 74 to X2 and its rest contacts 75 connected to the input wire 69 of its own stage and to a retaining wire 76 going from contact 66 of X2 to a winding of R3.



   The third and subsequent floors are a simple repetition of the second floor.



   Note that the counter of Figure 15 has the advantage that a carry pulse is transmitted through all stages in play without requiring the operation of the relays. For example if the three flip-flop circuits X are in position "1" (corresponding to 111 in the binary system or 7 in the decimal system), the closing of K sends a carry pulse directly through 66 of X1, rectifier 72, wire 71, contacts 66 and wire 76 of the second stage and by exactly similar connections from the third stage directly to the fourth stage. The flip-flop relays of the first three stages go together in position "0".



  Relay E is designed so that the flip-flop circuits cannot be operated until the restraint relays have taken their correct position.



  The end result in the case of this example is 1.000, sum of 111 and 1 in the binary system.



   Now consider the second floor for example.



   In the case of a single input, either coming from stage I through the closed contact "position 1" of the flip-flop circuit XI, or a po-
 EMI9.2
 tential applied by key 1, the IL relay operates since only one of its windings is energized. Through his work contact he makes
 EMI9.3
 operate the flip-flop circuit X2 If the flip-flop circuit Xz is in position "1", this operating potential is also transferred to the relay IL of the next stage and so on.



   If, on the contrary, there are two simultaneous inputs, a carry-over from the first stage coming from K1 and from contact 66 of the flip-flop circuit XI (in
 EMI9.4
 position "'Iie) and the other by% a direct potential by key E2. of the second stage, relay' Ç2 does not work since its two windings are energized and the potential is transmitted by the rest contact of relay IL to the rest at the IL relay of the next floor.



   FIG. 13 represents another exemplary embodiment in the-
 EMI9.5
 which the selection bars activate the input relays CI, G2, C3, 04, 05. In this embodiment the restraint relay IL has a single winding. Other restraint relays have two, and will operate if either or both are energized. The connections will not be described in detail as the operation is clear from the drawing, from the explanations given above. Let us take for example the flip-flop circuit XI in position "1" and suppose that the relays Cl and C2 are actuated by the bar-
 EMI9.6
 permutation res.

   Circuit XI is actuated by CI and relay p2 operates, the potential given by the NO contact of IL is not applied to binary relay X2 but is transferred by the two-way contact of C2 to relay R which operates by its first winding.
 EMI9.7
 working tact of f and the rest contact of C3's to-and-fro movement at rest, X3 is energized. Rectifiers are provided to prevent undesirable operation of the lower stage relays. It can thus be seen that the restraint relays can be actuated in two ways, namely when the circuit

  <Desc / Clms Page number 10>

   Previous flip-flop transmits a hold potential, ie when the previous hold relay is actuated and a direct input occurs simultaneously.

   Relay E in figure 13 has the same role as its correspondent in figure 12.



   When you want to make a subtraction, for example in the case of an error, it suffices to reverse the position of all the flip-flop circuits, to add the number to be subtracted and to reverse a second time every flip-flop circuits. This can be done easily with steps as shown in Fig. 14 because it suffices to energize an auxiliary winding 60 provided for this purpose. By energizing all of these auxiliary windings at the same time the position of all the stepping is reversed.



  Subtraction in the case of simple telephone relay flip-flops has a similar procedure and will be mentioned later in the description.



   General diagram of the electrical circuits - The circuits will be described with reference to the general diagram of figure 11. The apparatus represented by the rectangle 10 on the left side at the bottom of figure 11 comprises a machine to be written with a keyboard and the selection bars activated by the keys as well as the recorder intended to record the coded information corresponding to the characters. The system preferably comprises a recorder in which the encoded information already recorded can be read, decoded and transmitted to the flash photography device and to the variable exhaust of the film carriage while the next line is typed on the screen. keyboard. According to a usual technique, the different characters of a given font have widths expressed as a function of a common unit.

   For example in a typical font: the width of i and 1 is 5 units; that of f and j of 6; a, g and the digits of 9; b, h, n and s of 10; J, F and L of 12, M, H and W of 15. An arbitrary value is also assigned to the minimum space between words which can for example be 4 units.



   By pressing the key corresponding to a character, the recorder is activated and records the corresponding code. At the same time, for each keystroke of a character its width is transmitted to the line counter 12. For this purpose, four conductors connect the rectangle 10 to the rectangle 12 (figure 11). It will be noted from the foregoing that twelve different widths, ranging from 4 (minimum space) to W (15 units) are generally sufficient for all possible castings For this, four threads 14 are sufficient, given that with four threads there are 2 = 16 different combinations, although if more width values were desirable, more threads could be used.



   Lowering the space bar (SP), besides adding 4 units to the row counter, adds 1 to the spaces or CI interval counter, so that the row counter accumulates the number of units from the beginning of the line and the space counter the number of justifiable spaces. For this, a connection brings together a contact actuated by the space bar of the typewriter at the counter CI.



   In addition to the wires 14 connecting the selector bars to the line counter, there is a conductor 20 excited by a so-called "universal" contact of the typewriter which operates when any key is depressed and whose object will appear later. .



   At the end of the typing of a line, the operator reads again and if he finds the correct line, he activates the send key 22 which starts the justification mechanism and then the printing automatically.



  The rationale will now be described briefly.



   Let J be the desired justification expressed in the chosen unit of length and L be the total of all the widths of the various characters and the minimum justifiable space as recorded in the recorder.



  The line deficit is D such that J - L = D.



  The object of justification is to distribute between justifiable spaces

  <Desc / Clms Page number 11>

 the deficit D such that the length of the line is increased from L to exactly J.



   Let N be the number of justifiable spaces, the relation between N and D is expressed by: D = N Q + R (1) where Q is the integer part of the quotient of N by D and R the remainder. If the quantity Q + R / N were added to each space, it would be necessary for the cart to be able to move by any fraction of a unit. Such devices have the disadvantage of being delicate to adjust and that errors are added.



  In contrast, in variable escapes which move only in whole numbers of units, the errors do not add up. To benefit from the advantages of this type of variable exhaust for the film carrier, the justifier adds Q units to the N- R first spaces and Q + 1 units to the last Rs. The number of units thus added is: (N - R) Q + R (Q + 1) = NQ + R = D (2)
The operation of the justifier will be described in relation to a digital example. Suppose that D is equal to 41 (that is, the line is too short by 41 units) and the number of spaces N is equal to 12 (13 word line).



   The number N is sent to the counter 12 several times in succession. This is done by the 5 conductors 23 connecting the space counter CI with the line counter 12. In the example chosen, the deficit D is consequently reduced from 41 to 29, then to 17 and to 5. At the fourth addition, the capacity of the counter is exceeded (by 7 units in the example chosen) o The number of times that N has been added in the counter to make it "switch" is equal to Q + 1. If we write the general equation (2) again and below the values of the numerical example we have: (N - R) Q + R (Q + 1) = NQ + R = D (7 x 3) + (5 x 4) 12x3 + 5 = 41
A "rest" switch PR is connected to counter 12 by a set of 5 conductors 24.

   The counter, as already noted, at this time indicates 7 (or in general N - R) o This value of N - R is transmitted by the wires 24 to control the installation of the switch PR.



   Also provided is a so-called "zone" switch SZ connected to the space counter CI by a conductor 25. This zone switch first registers the number Q which must be added to the first N - R spaces.



  In the example described the switch SZ selects an increment of 3 units which is to be applied to the first 7 spaces. A conductor 26 connects the remainder switch PR to the zone switch SZ by which PR sends a pulse to SZ when it exhausted N - R by advancing from N - R not. In other words, when the carriage exhaust is operating at print time, the remainder switch PR advances one step to its home position for each justifiable gap. On reaching its rest position, that is to say when the first N - R spaces have been increased by Q, the pulse sent by PR causes the switch SZ to switch from zone Q to Q + 1, whereby the remaining R spaces will be increased by Q + 1 units.

   (In the example chosen, PR must take 7 steps to reach its home position and then change the zone switch from 3 to 4 units for the 5 remaining spaces.



   Referring again to the diagram of Figure 11, rectangle 28 represents the width translator, signal translator and repeater.



  This rectangle 28 is connected to 10 by two sets of wires, represented by a set 30 of 4 width indicator conductors and a set 32 of 3 character selection conductors in the same width. These conductors are excited by the readout mechanism of the recorder. The 4 threads 30 serve to select the width of the character and to ensure correct operation of the variable exhaust. The 3 wires 32 select the desired character among those having

  <Desc / Clms Page number 12>

 the same width. For example b and h have the same width, so that the wires 30 will transmit for the two identical signals. However, the wires 32 send different pulses. If there are no more than 8 characters of the same width, the 3 threads 32 are sufficient.

   In fact, in most slots there are more than 8 characters of certain widths. This can be taken into account by increasing the number of wires 32. Such additional conductors are not necessary because the wires 30 are not fully utilized and some unused widths can be allocated to others, as will be seen later. . From tractor width 28, a 12 conductor cable 34 goes to variable exhaust 36. These wire pulses 30 and 32 are also transmitted through a signal translator and a 7 wire repeater to the photographic unit 40. , in which the selected character is illuminated as it moves to the photographic position to project an image onto the film.



   The zone selector SZ is connected by seventeen wires 42 to the rectangle 36. In the event that a space signal appears in wires 30 and 32, the flash circuit is not actuated but a signal is sent by the signal translator to the zone selector SZ by a wire 43. The width of the gap is determined by the position of the zone selector, and a signal is transmitted to the variable exhaust by one of the wires 42. The width minimum of a justifiable space is 4 units and given that in practice it cannot be given a width greater than 18, the threads 42 are sufficient. Spaces 0, 1, 2 and 3 units are not used and these values are reserved for justification between characters, as we will see later.

   In the example chosen the zone selector will give by the variable escapement a justification increment of 3 units for the first seven spaces and an increment of 4 units for the last five so that the width of the first seven spaces will be 7 units and 8 units for the last five.



   The mathematical principles on which the justification is based were stated as if the calculations were made in the decimal system. However, the actual calculations are preferably done in the binary system. The components used and the method of calculation will now be described in detail. However, since an important element of the system is a special counter, the principles of the operation of the counter will be explained first, after which the various parts of the system will be described in detail.



   Line Length Counting - The operation of the line counter will be described in detail in connection with the detailed circuit shown in Figures 16 to 21. The typewriter is shown at the bottom left of Figure 18.



   When the operator lowers a key, he activates a number of permutation bars and closes the associated contacts cdl, cd2, cd3, cd4 which characterize the width of the characters and the contacts cd5, cd6 and cd7 which differentiate the characters of the same width and do not intervene in the justification. Since the contacts of the changeover bars can be closed for slightly different times, intermediate relays R11, R12, R13 and R14 are inserted between the changeover bars and the counter. These relays are actuated by the wires 14 and the contacts of the permutation bars and are blocked by holding circuits established by their closing contacts before breaking and a closed working contact, at this time, of a special delay relay. DR.

   In this way it is possible to keep these relays actuated for a sufficient time regardless of how long the selector bar contacts are closed., The intermediate relays, when actuated, apply a battery to the input wires. counter through a cs switch relay to zero. The counter (figure 19) comprises 5 stages with 5 restraining relays R2 to R6 and 5 combinations of binary relays A1 - Bl to A5 - B5 of the type shown in figure 15 The counter is identical to that of figure 15, the combinations of relays Al - Bl, A2 - B2 being the same as Xl, X2, etc. in figure 15.

  <Desc / Clms Page number 13>

 



  Four stages are sufficient for the widths of the characters, but it is desirable to provide a fifth stage in order to be able to have more than 15 justifiable spaces since, as previously stated, the number of spaces N is sent into the counter by the wires 23. Whenever any key on the keypad CL is pressed, a contact u, called a universal contact, is actuated. The contact activates a relay RU which also locks on a closed contact of the DR relay. Another relay E is actuated by the working contact of relays RI-1 to RI-4 and RU in parallel. The purpose of this relay E is to delay the application of the battery to the binary relays and thus to give the restraining relays R2 to R6 time to operate and to prepare the circuit.



   When the binary counter exceeds its capacity, a pulse is sent to the ACS (Accumulation Stepper) stepper which advances one step.



  The ACS switch is a step-by-step (Figure 18) which advances one step each time the binary counter counts 32. It can be thought of as a base counter 32 with sufficient capacity to record the number to - number of units of the longest line. The pulse is transmitted in the following way: When the binary counter toggles, a circuit is established through the carry relay R6 in the same way that the holds are made in any stage of the counter itself, as explained in relation to figure 15.

   When the relay R6 closes its contacts, it transmits a pulse through wire 80, a contact of the COS switch which will be described later and a wire 82 to ACS. The excitation of the winding followed by the cut when R6 falls again pushes the ACS brushes forward by one step.



   It has been noticed that the RU relay is used for the case where there are more than 8 characters of certain widths, so in a given font there are 20 characters having 9 units of width and 14 of 10 units. Since widths 0, 1, 2, and 3 are not used in a normal font (the narrowest sign being the minimum space of 4 units), these unused values can be used to increase the number of characters. teres of certain widths. This is done satisfactorily by the
 EMI13.1
 circuit shown in figure 19.

   The relay BU which is actuated for all characters only comes into play when the binary values 0, 1, 10 (two) and 11 (three) appear on the relays Rll and R12. In other words, every time that R13 and / or R14 are not actuated A connection goes from the battery through the make contact of RU, the rest contacts of R14 and R13, a back and forth from R12 to the connection point of the two rectifiers rd3 and rd4, respectively connected to the input wires of the counter having the values of 1 and 8 (1000 in binary). Thus when no relay RI is actuated a signal of value 9 is sent through the two rectifiers in the counter. When relay R11 is only actuated, the same value is sent to the counter, rectifier rd3 then being shunted by the work contact of RI 1.

   When the counter RI 2 is only actuated, its normal value of two is brought to ten by the two-way switch in the actuated position and connected to the fourth input wire of the counter. Finally, it is clear that RI I and RI 2 are both activated and a value of eleven units is sent to the counter.,
Representation of combiners and switches - For some switches, in particular ACS and SZ, the contacts have been represented in isolation. For others, however, a usual representation in telephony has been adopted. These different levels (or cams) are designated by capital letters A, B, C, etc. Thus the COS correction combiner (figure 18) has 7 levels designated by the letters A to G.

   The contacts of these different are designated by:
 EMI13.2
 AeOVOGCOee0 1/4 Gaaaaam.aaamam. 3 D & oeoeeQ & 006 0 Ee0CGQG.O.000 4 F a o o e o 0 0 o a a a o e 3

  <Desc / Clms Page number 14>

 
G ............ 2/4
H ............ 2/3 Each number corresponds to a position where the contact is closed. The sign means that the contact is closed between the two positions, while the dot means that the contact is open between the two positions. Thus level A is open in its rest position 0 but closed for all positions ranging from 1 to 4. Level C is closed in position 3 only, D in position 0 only, E in position 4, F in position 3, G on 2 and 4 but open in position 3 and H on positions 2 and 3.



   Correction of errors - Before describing the justifiable space counter and associated devices, the procedure used to correct errors will be explained. If the operator notices that he has made a mistake, it is necessary to erase the erroneous character of the recorder and to subtract its width from the line counter, before replacing it with the character correct.



   To subtract from the counter the width of the erroneous character without having to retype the latter, it is necessary to detect the combination corresponding to this character. To this end, after having brought back the carriage of the typewriter so that the character to be erased is in the typing window, one excites, in the manner which will be described later, the GOG electromagnet which attracts its Y3 armature. by rotating it anti-clockwise around pivot Y4. This frame pushes the plate OP upwards (FIG. 18) against the action of the spring Y2, this movement being guided by the studs Y6 cooperating with the buttonholes of the OP.



  This OP plate is fitted with 4 small Y levers that can each pivot around an axis mounted on the OP. The right-hand ends of these levers are found under the rows of pins on which the line is recorded and their left-hand ends cooperate with 4 Y8 correction contacts.



  When the OP plate gets to its upper position, it is clear from the figure that the Y levers will have encountered the pins that were not depressed, which will force them to open their corresponding contact. Thus in the example shown in figure 18, it has been assumed that the plate OP is slightly pushed upwards and it can be seen that all the Y levers, except the third from the top, have encountered a non-pin. pressed, so that only the third contact Y8 from the top will remain closed, which means that only pin N 3 has been inserted among the first 4 pins, and that the character to be deleted is 4 units wide .



  The device described therefore makes it possible to detect the combination of a character to be erased in order to subtract it from the counter. It is then necessary to erase the bad character combination from the recorder, that is to say to return to the rest position all the pins of the recorder located on the vertical row where the character to be erased was checked in.



  This result is obtained by exciting the electro EF in the way which will be described later. This electro attracts to the left (figure 18) and against the action of the springs the Y5 plate which can slide on the Y7 studs. This plate carries hooks such as CF which meet the jaws B C (figure 2) of the depressed pins and thus return to the rest position all the pins of the vertical row of the recorder where the deleted character was located.



   The correction is done automatically.



  After returning the character to be erased to the typing position, the operator presses the COK correction key. This activates the COS correction switch which determines the sequence of the correction operations. Subtracting the width of the character is done by inverting the binary counter (i.e. changing all I's to 0's and 0's to I's), adding the number to be subtracted, and inverting a second times the binary counter.



  This is the mathematical process of subtraction by addition of complements.



  For example, to subtract II from 1000 in a binary system, we can invert I000 which gives OIII, add II to obtain 1010 and invert a second time to obtain the result 0101. This is done by COS which, when

  <Desc / Clms Page number 15>

 COK is released, falls back and is connected by its level A to a source of pulses G comprising a cam Ca2 (FIG. 16) mounted on the camshaft of the photographic assembly. The COS switch thus moves 4 steps and finally falls back to position 5 which is also its starting position and on which it is disconnected from the source of pulses.

   While moving, this switch performs the following operations: First of all, it disconnects the ACS winding from the working contact of R6 by its level D-0 (note that level D does not give a con - closed tact only at position 0). In position 2, by its level G, the switch COS sends an impulse via wire 84 to the INV relay (inversion relay). This relay closes its contacts and sends a pulse through each of the wires 120 to all the relays A1 to A and B to B of the binary counter.

   By referring to the description of the counter, given previously, it will be seen that this, when the relay INV drops, will invert all the binary stages, that is to say it will pass all the stages which are to position I from this position to position 0 and vice versa.



   At position 3, COS level C connects the SUS switch (subtraction switch) to the working contact of relay R6 through wires 80 and 85. In other words, it disconnects ACS from restraint relay R6 and connects SUS in its place for reasons which will be explained later. In positions 2 and 3, COS operates (by wires 86 and its level H) the COC probing electro which detects the position of the first 4 pins of the recorder of the initial row on which the character to be deleted is recorded and thus determines the width of the character to be erased, as explained above.



   In position 3t and by its level E, COS sends an impulse via wire 87 to the correction contacts of the recorder. These correction contacts which remained closed (see figure 18) correspond to a pressed pin of the recorder. The Y8 correction contacts are connected to the meter by the COR1 .... COR4 wires on which RY switchers are mounted to avoid mixing of circuits. In the case of the figure, the third contact starting from the top remains closed, and the impulse arriving by the wire 87 will be directed on the relay R13 as had occurred at the time of the striking of the erased character.

   In position 3, a pulse is also sent on the universal wire 20 by the level E of COS (wire 88) or by any other means in order to achieve exactly the same conditions as for typing the wrong character.
Y9 switchers prevent hammer operation during correction.



   If a carry between the meter and the accumulator switch occurs during the correction, it is necessary to subtract it from the ACS switch since this means that this carry would occur if the false character were added at this time instead of subtract it, if it has not yet been saved. Since the commercial step-by-step switches are generally not suitable for reverse gear, an SUS correction switch is provided. This switch receives all carry from the counter that appear during corrections, and moves the "end of line" terminal of the ACS switch back one step for each carry.

   We can see in figure 18 that ACS gives a signal when its b1 brush arrives on terminal f, which takes place when the capacity of the counter is reached or exceeded. But we can see that if SUS has moved by a step, ACS will have one more step to do In position 4, COS, by its level F ;, sends a second pulse on the INV relay which again inverts the position of all the stages of the binary counter. The width of the character to be deleted has thus been subtracted from the line counter. In position 4, by its level D, the COS switch activates the electro EF which resets all the spindles of the row on which the character to be deleted had been recorded.



  At the end of the fourth pulse, the COS switch returns to position 5 (which is also the home or zero position) and disconnects from the pulse source. The erroneous character was thus deleted at the same time

  <Desc / Clms Page number 16>

 on the recorder and on the meter and the operator can then replace it with the correct character.



   Justifiable Space Counter - While the line length is recorded in binary numbers in the line counter, the CI justified space (or interval) counter records the number of spaces (Figure 20). This counter CI is also a step-by-step type switch and receives a pulse whenever the space bar SB is actuated by the operator. It is provided with 5 levels A to E which translate the number of spaces into a binary number to be used in the justification calculation and has a capacity of 25 points, sufficient in practice.



   According to the convention mentioned above, the different levels of CI establish their contacts as follows:
AT . 1.3.5.7.9.11.13.15.17.19.21.23
B 2/3. 6/7. 10/11. 14/15. 18/19. 22/23
C 4/7. 12/15. 20/23
D 8/15. 23/24
E 15/24
Z 1/24 (return to zero)
Calculation of Justification - The operation of the justification will be described in relation to a numerical example. Suppose for example that the composite line is too short by 41 units, or in other words that the deficit of the line is equal to 41, and that there are 13 words and therefore 12 justifiable spaces.

   In this case, the binary relays of the line counter are all in position "1" except the relays A4-B4 (8 units) and the ACS accumulator stepping one step from its "f" position (32 units). Forty pulses are necessary to bring all the binary relays to position "1" and to bring ACS to position f plus one pulse to send a carry signal and return all the binary relays to 0 "denoting as well as the capacity of the device. counter was exceeded, for a total of 41 pulses.



   The number of spaces N was recorded in the CI counter which advanced 12 steps under the control of the typewriter space bar. It is therefore in position 12.



   The levels A to E of the space counter can be thought of as a converter to transform the number of steps into binary numbers. Thus a closed contact on level A represents 1; on level B, 2; on C, 4; on D, 8 and on E, 16. Thus in the example chosen we see that in position 12, the contacts of levels C and D are closed, corresponding respectively to 4 and 8, that is to say to 12 in the decimal system.



   The succession of operations brought into play by the justification is controlled by a step-by-step combiner FLK. This FLK switch has the contacts of its different levels set as fired:
A 1
B 4/5
C 2
D 2
E 3
F 5
G 4
H 5
1 2
J 0/3 .5
K 2
When the operator lowers the sending key 22, the FLK winding is energized and when the key is released FLK goes to position 1, where by its level A its winding is connected by wire 90 to a source of

  <Desc / Clms Page number 17>

 G pulses, giving an interrupted current at the rate of 10 per second
 EMI17.1
 for example. FLK immediately switches to position 2 where it stops.

   In this position, it actuates the switch relay CS which connects the 5 levels A to E of CI to the line counter and through its level C, FLK directs the pulses G to the common point 94 of these five binary levels of the counter of CI spaces.



  As explained above this has the effect of adding 12 to the row counter for each pulse. The deficit D is consequently successively reduced from 41 to 29, 17 and 5. At the fourth pulse, the capacity of the counter is exceeded and the restraint relay R6 is actuated in the manner which has been described previously.

   A pulse is then transmitted through wires 80 and 100 to level D of ACS which is at this time on pin F, which represents the position of “full line” or “end of line.” Note that if any Corrections were made which resulted in an advance of SUS, the full line will be represented by a right contact of 1. of an equal number of steps.
 EMI17.2
 Zion propagates through a contact from level D of SUS, wire loe, level D of FLR to the coil of FLK. At the end of the pulse FLK changes to po-
 EMI17.3
 sition 3 and by opening the contacts of its level C stops the sending of pulses in the line counter through CI.

   At this time, the line counter has switched !! and records the difference between 12 and 5, that is to say 7. While FLK was in position 2, the zone switch SZ had its coil connected to the common point 94 of levels A to E of CI by the thread 25 and
 EMI17.4
 FLK level C. The selector SZ has therefore advanced by Q + 1 step, that is to say in the example chosen by 4 steps and is in position 4.



   In FLK position 3, two things happen. First, CI is returned to its rest position, since it is no longer needed, by means of its Z return level which is closed for all CI except the zero position, the contact rz of CI, the wire 104, the level E of FLK in the manner well known in advance. Second, the remainder recorder takes place under the control of the row counter.



  It has been explained how the row counter recorded N - R, 7 in the digital example described. This means that the binary relays A1, A2 and A3
 EMI17.5
 are in their "II" position. These relays, through their working contacts, establish a bell-like operating circuit for PR.

   This PR switch is a step-by-step switch whose various levels make the following contacts
 EMI17.6
 A fel.j.eVaoloool4.aJ.0al0 20. 22. 24 B 2/3. 6/7 10/11. 14/15. 18/19. 22/23
 EMI17.7
 C 2/5 o 10/13 18/21
D 1.10/17
E 1/9 F 24
Z 1/24 (Return to zero) The bell operating circuit of P goes from the battery through
 EMI17.8
 level E of FLK (in position 3), wire 106, a normally closed contact of an NJ relay (which will be described later), wire 108 at the common point of levels A to E of PR, wires 24 line counter flip-flop relays.



  It can be seen that this circuit is closed until the moment when PR reaches the posi-
 EMI17.9
 tion 18 (in the numerical example chosen) where it stops, ie it stops 7 steps from its neutral position 25. During these two operations, the FLK winding was kept energized. through rectifiers 110 and 112. When CI and PR have both reached their respective positions as described above, the battery is disconnected from FLIC, which then drops to position 4.

   where by its level B its winding is again linked to the source of pulses G, so that it immediately passes to position 6, which is also its rest position o In position 4 of FLK, the binary relays of which the battery wire passes through level J of the dropped FLK and in position 4, FLK operates the relay JR by its level G and wire 114, the role of which will be explained later.

  <Desc / Clms Page number 18>

 



   In position 5 of FLK an impulse is sent by the level H of FLK and the wire 116 to 1-electromagnet EM which controls the exhaust of the carriage of the typewriter so as to place it in front of the first row of pins of the recorder, that is to say in position to command the typing of the next row. Indeed, with the recorder with two rows of pins shown in Figure 2, it is necessary to return the carriage of the typewriter to a point where it is clear of the pins of the recorder to allow switching from one row of pins to the next by moving the pin field vertically.



   The justification is then prepared and the machine ready for photographic composition of the line while a new line is recorded.
Lines not justified - If the operator wishes not to justify a line, he lowers the NJK key before activating the send key 22. This operates the NJ relay and locks it on level B of FLR. The NJ relay therefore remains actuated for the duration of the justification computation and during the photography of the line and a little longer under the same conditions as the SE relay. In position 2 of FLK, the relay NJ sends a pulse to the winding of FLK so that the latter does not stop at this position and only sends a pulse through CI.

   In position 3 of FLK, relay NJ opens wires 106 and 108 and thus prevents PR from operating. CI only, works in this position 3 of FLK (through wire 104) and returns to zero.



  FLK goes to positions 4 and 5 as previously described. SZ has received a single pulse and is therefore in position "1" and selects piston 4 of the variable exhaust. All justifiable spaces are given the same width of 4 units.



   Deletion of a line - If the operator wishes to delete a line, he activates the KLK key which blocks the NJ and KL relays before activating the send key 22. The justification calculation does not take place as explained below. high and the CR cart relay is prevented from operating by an actuated KL make contact which cuts the battery of wire 134. The push-pull of the KL relay supplies battery to the portion of wire 120 that goes to contact w of the translator circuit to allow the "end of line" signal to agiro A rest contact of CR at rest disables the camera circuit via wire 136 which bypasses the decoder. When the reader of the recorder encounters the signal "end of line", a battery is applied by the signal translator to the winding of FLR by a working contact of KL.

   The film advance device is turned off by a KL rest contact on wire 138. FLR makes one full turn and keeps the recorder explorer SE relay actuated long enough to clear the picture completely. line.



   By moving back and forth on the NJ relay, a different level of FLR can be selected for the advance of the film after an unjustified line, i.e. automatically changing the spacing. between lines at the end of a paragraph.



   Adjustment of the justification - The JR justification relay (figure 19) was mentioned during the description. This relay has 5 sets of contacts 118 in series with 5 manually operated keys JK1. The contacts 118 are connected to the inputs of the flip-flop circuits of the row counter by the wires 120. If none of the JK1 keys are closed, the counter will remain empty when the relay JR is actuated. If any of the keys is closed, an initial value will be recorded before striking the line when FLK sends a pulse to position 4. This reduces the available capacity of the counter accordingly.

   If, for example, the total capacity of the line counting system (12 and ACS) is, for example, 512 and if you want a line of only 504 units in length, 8 will be sent to the binary counter at the start of each line, keeping the corresponding key JK1 closed o If the length of the line must be reduced by more than 31 units below the capacity of the counter, the initial position will be modified by moving a bar of a key JK2 (figure 18) which closes all contacts, except

  <Desc / Clms Page number 19>

 one, the rollback level of the ACS paa-stepper.



   Photographic composition - When the line is ready to be photographed, a bs contact (figure 17) is temporarily opened and the RS relay (figure 17) which was at rest, because both ends of its windings were connected to the battery, is working and hangs on resistance rb. The RS relay prepares the operating circuit of the CR relay which operates when the start cam Ca7 closes its contact at an appropriate time in the cycle. The relay SE which controls the operation of the explorer trolley EXP of the recorder, also operates by the same circuit as CR but is blocked on the level B of a switch FLR (figure 20) whose contacts are shown in the drawing. This switch, also of the step-by-step type, controls various auxiliary operations such as film advance.

   The CR relay completes the operating circuit of the variable exhaust of the film carrier which comprises an electro-motor E and eighteen solenoids which determine the number of escaped teeth and subsequently the advance of the film. film, Simultaneously CR applies a battery to wire 119 by a contact, normally closed, of a relay KL described in connection with the half-line erasure o The wire goes from contact w of the width translator circuit comprising the relay Tl to T7 (figure 9). The contact w is the first of a back-and-forth set constituting the fan-shaped circuit 121 of the relays T. By the wires 30 and 32, these relays are actuated by the contacts of the reader or sensor of the recorder C1 to C7 shown schematically in figure 18.

   When the relay SE operates, it closes the circuit of the electric EXP which controls the exhaust of the exploration carriage via the wires 122 and the cam Ca8. The relay SE closes the circuit of the electro PLP by the wires 124 and the cam Ca9. The electro PLP pushes up the plate 104E and the read contacts which correspond to the pins inserted during the strike of the line are closed for a moment. The corresponding relays T operate and are blocked through a blocking circuit comprising the Gala cam, a wire 126 and the contacts of the CR relay.

   These T relays have several purposes, firstly, they select the character to be photographed by their rest contacts 128 which control the operation of the decoder 130 and the lightning circuit 40 by the wires 38, secondly, by their va- contacts. and back and forth 121, they choose a variable exhaust piston through the wires 34 and thus control the movement of the film carrier according to the width assigned to each character. When, for example, pins 1, 3 and 4 of recorder are in their depressed positions, relays T, T and T are actuated and the wire selected is that of the stop piston l 'since the binary value of the pin combination 1-3-4 is 1 + 4 + 8,

       that is to say 13. Finally, thirdly, the relays T translate certain signals such as "justifiable space", "end of line", "justification between characters" and "change of font" by the wires ji, f1 , jc and cf.



   When the read contacts find the combination "3" in the recorder, relay T alone is actuated and the potential directed by wire 43 to the brush of SZ which is, in the digital example chosen, in position 4. The pins of SZ are connected to the stop pistons of the variable exhaust as follows: pin 0 and 1 to piston 4, pin 2 to piston 5, pin 3 to piston 6, pin 4 to piston 7 and so on. In the present example, SZ being in its position 4 selects the piston 7 and the variable exhaust escapes 7 teeth and advances the film carrier by 7 units for each space. At the same time PR also advances by one step for each space being also connected to wire 43.

   A wire 130 connected to 43 passes through the level F of PR and a rest contact of NJ towards the wire 25. When PR reaches its position 24, it sends by its level F a pulse on the wire 130 towards SZ which advances. of one, no. Since PR was 7 steps (usually N - R steps) from position 24, the first spaces will have a value of 7 units. The following R or 5 spaces will have a value of 8 units. This makes a total of 7 7-unit spaces, or 49 plus 5 8-unit spaces, or 40 units and a total of 89. Since the counter has recorded 4 units for each space, the. line length has been increased by 89 -

  <Desc / Clms Page number 20>

   48 only or 41 units.

   Given that. this was the value of the deficit in the example chosen, the line is justified.



   At the end of the line, the read contacts find on the recorder the combination 3 -.5 that the operator has recorded by actuating key 22. Relays T3 and T5 are actuated and direct the battery to the Common point. of resistor rb and relay RS. The RS and CR relays then have both ends of their windings to the battery and drop out.



  The SE relay, on the other hand, remains actuated for a certain time by level B of FLR, because it is necessary to move the explorer carriage a few more steps to replace the last pins of the recorder.



  When CR releases, it applies a battery via one of its rest contacts to the clutch DEB which normally connects the variable exhaust EV to the film carrier. The carriage is thus released from the variable exhaust and returns to its rest position under the action of its return springs. When it reaches its starting position, the truck activates the RT contacts. This disconnects the so-called clutch battery and the film carrier is re-engaged to the variable exhaust. On the other hand, the battery which was applied to the winding of the FLR stepper is cut by operation contacts RT so that FLR advances by one step, its winding is connected by its level A to the source of pulses G and it begins to rotate.

   On closing, the RT work contact connected the variable exhaust motor EV to the battery through the cam Ca6 and energized the piston 1 of the variable exhaust. The film carrier which, under the effect of its inertia, has exceeded its exact starting position, starts up step by step until the circuit is cut by the MP cam which has a corresponding isolated position. - dant to the exact starting position. The FLR stepper, which keeps the SE relay actuated, also sends pulses to the film advance device FF which actuates the debtor driving the film. The number of pulses sent can be changed by means of a multi-position switch FFK, each position of which selects an FLR level, each of these levels sending a different number of pulses.



   Flash Circuit - The flash circuit (Figure 16) operates as follows:
The photographic unit comprises a disc D on the periphery of which are the characters CY, for example transparent. On an opaque background, an objective 0 projecting an image of these characters on the sensitive film F and a flash tube L. The disc rotates continuously and the characters are photographed by sending an electrical impulse through the discharge tube which lights up for an exceedingly brief duration, a few millionths of a second. Despite the high linear speed of the characters, perfectly clear images are obtained.

   To ensure character placement on film, the instant at which the flash occurs must be determined with extreme precision, also in the order of a millionth of a second. This is obtained by means of a photoelectric pulse generated by a slit F associated with each character. Since it is possible to arrange a slot corresponding to a given character with absolute precision, for example by photographing them simultaneously., The precision does not depend on the position of the characters on the disc but only on the relative position of the character and the slot associated with it.



  The photoelectric pulse is obtained by means of a film reader P, of the type commonly used in talking movies, which projects a very fine light slit on the path of the slits F and on the photoelectric cell PH. The photoelectric pulses which occur at the rate of one for each character are inoperative as long as the decoder (figure 10) has not opened to them -a passage to activate the flash lamp L.



   The decoder 142 generates at the moment when the character passes into the shooting position a relatively large pulse which acts on a window lamp and lets the corresponding photoelectric pulse pass. The pulses from the PH photocell are amplified by a pentode lamp 144 and appear with positive polarity in the cir-

  <Desc / Clms Page number 21>

 anode fired.

   They are then applied to the control grid of the window lamp 145, also a pentode with a suitable bias potential supplied by a battery 146. This lamp 145 is normally blocked because its screen grid is connected to the battery. mass by a resistance 147.

     On the other hand, the decoder circuit comprises an inductor 148 and a battery 1490 As long as the decoder circuit is closed, a current of a few milliamperes limited by the resistor 150 circulates in this inductance. of the decoder find in the decoder the same combination as that of the recorder, the circuit is opened as explained in relation to figure II and an induced breaking current appears across the choke. The amplitude and the shape of this pulse depend on the resonant circuit formed by the inductor 148 and the capacitor 151.

   This induced voltage, of positive polarity, is applied to the screen grid of window lamp 145 and makes it conductive and amplifying while the character passes in front of the lens. At the precise instant when the character is in the shooting position, the photoelectric pulse occurs, passes through the lamp 145 and is transmitted to the lamp 152 which reverses its polarity and limits its amplitude. . The resulting pulse of positive polarity is sent to the lamp 153, preferably of the commercial Sylvania CAS type, and a capacitor 154 is discharged, through this lamp 153 thus started in the induction coil 155 of the type used in the ignition system of automobiles.

   This discharge generates a pulse of several thousand volts which ignites the discharge lamp L and the capacitor 143 is discharged causing a flash of light.



   To prevent the discharge lamp from operating more than once per revolution, capacitors 143 and 154 are disconnected from the potential source at the start of the decoder active cycle by a Ca3 cam. These capacitors are charged through appropriate resistances in the part of the next cycle attributed to the movement of the film carrier.



   Time control - The photograph of a character consists of two phases or cycles, a photographic cycle during which the character is photographed and a forward cycle during which the film carriage moves, The cams controlling these operations are shown on the fig-- re 16 and are driven by the same shaft as the die-carrier disc.



  Each opens and closes a contact in 1 - 1 order and for the relative times shown in the diagram of Figure 22. The disk carries the die characters on half of its periphery. The two successive cycles occur during a single revolution of the disc. There are 10 cams, the Cal cam controls the return of the film carrier to its precise starting position, the Ca2 cam is a pulse generator (source G) used in the justification, the Ca3 cam bypasses the decoder during the feed cycle and prevents the production of flashes of pictures, the Ca4 cam charges the capacitor 143 of the discharge lamp and prevents more than one flash per revolution from occurring, the Ca5 cam controls the operation of the variable exhaust advance selection pistons,

   the Ca6 cam controls the operation of the electro-motor of the variable exhaust EV, the Ca7 cam is the start-up cam which starts printing by closing the operating circuit of the CR and SE relays at the appropriate moment, the Ca9 cam controls the PLP contacts of the explorer (probes) and the calo cam closes the blocking circuit 126 of the relays Tl to T7 so as to release the contacts of the recorder as soon as the relays T are blocked for the rest of the cycle.



   Changing the Font - Multiple die fonts can be placed on a disc, for example one on the first half of the disc periphery and another on the other half. As discussed above, the photograph occurs during the first half-revolution and the carriage advance during the other half. The change from one cast iron to another requires in the operation of some cams an offset of half a tôura This can be done, for example by the ACFe relay Each cam from Ca3 to Ca10 has two sets of contacts offset by 1 / 2 turn and connected to the central blade of the ACF relay back and forth. When this relay operates, it transfers the cir-

  <Desc / Clms Page number 22>

 fired from one cam contact to another.

   As a result, the shooting cycle takes the place of the feed cycle and vice versa.



   The ACF relay is controlled from the recorder. When the feeler contacts find the combination 3-6 (corresponding to the cast iron change signal) in the recorder, a potential is applied by relays T3-T6 actuated to the ACF and BCF relays (figures 16 and 17). The ACF and BCF relays and the rc and rd resistors constitute a flip-flop circuit. At the same time, the control potential is applied to the "zero" piston to prevent the operation of the variable exhaust during the change of cast iron. When the cast change pulse is complete, the flip-flop circuit switches to its second position and remains there until the explorer finds the 3-6 combination in the recorder, whereupon the machine switches back to first cast.



   Mechanical counter - Figures 23, 24 and 25 show an exemplary embodiment of a mechanical binary counter which can replace the electro-mechanical counters described above. This counter essentially comprises double-entry epicyclic gears interconnected by transfer members transmitting an elementary displacement for two elementary displacements received. The counter shown by way of example comprises five trains, four of which have double entry. The indices of the components of the train counting "1" are followed by A; the components of the train counting "2" are followed by B; the components of the train counting "4" are followed by C, etc ...



  Each train is made up of IM and 2M planets and 3M satellites. The inputs are made by the planet carrier arms 10M and the transmission of the restraints by the planetary 8B. The sun gear 2M is integral with a sleeve 7M on which is mounted the transfer wheel 6B which receives the retainers from the lower stage, that of the units in the case of the figure. The planetarium 1M is integral with the toothed wheel 8M and transmits the restraints to the upper stage. The entries are made by the ratchet wheels 5A .... 5E wedged on the axes 4A .... 4E which are integral with the axes 10M on which all the satellites such as 3M are mounted.

   The operation is as follows: three cases are to be considered. 1 Simple restraint: a restraint coming from the lower stage is transmitted by the transfer members 9A, T and 6B and causes an elementary rotation # of the planetary 2M; the transfer pinion 9B advances one step #. Simple entry: it is made by turning the ratchet wheel 5B by angle "U" = # and the transfer pinion 9B in this case also turns by angle 2 in the same direction as in the case previous. Simultaneous Entry and Hold, in this case the transfer pinion 9B turns two steps, ie by angle 2 ci) in the same direction as in the previous example.

   Each epicyclic train causing a reversal of the direction of rotation, the direction in which the ratchet wheels 5A, 5B, etc. are rotated, alternate from one train to the next. The pairs of gears 8A-9A and the transfer pinions T-6B are chosen in such a way that the successive elementary angular displacements of the planets IM are transmitted only once in two in absolute value and in sign to the planetary of the retainers of the upper floor, a Maltese cross or derivative system preferably being used to prevent the transmission from one floor to the other of the sets and to block each train in the rest position.

   In the example of the figure, the ratchet wheels 5A .... 5E have 20 teeth and are actuated by a pawl and an electromagnet shown schematically in figure 25.



  An angular displacement "U" or 2 W therefore corresponds to 1 / 20th of a turn or 18 degrees. We have seen that the elementary displacement of the planets is plus or minus 2U, that is to say 36 degrees, alternately in one direction and the other, so that the elementary displacements coming from a reservoir or an entry add up. Each transfer element consists of two toothed gears 8A and 9A and a transfer torque T-6B. The ratio of gears 8-9 is such that each elementary rotation # of planetary 8A corresponds to a 90 degree rotation of pinion T.

   This pinion has only two teeth 27M and 28M in the same level T6 and its shape is such that, combined with the profile of the wheel 6A, the latter is locked by the hollow profile 20M until a tooth 27M hole-

  <Desc / Clms Page number 23>

 ve a hollow 19M and drives this wheel 6A until the tooth emerges from the hollow, the wheel being locked immediately. The SA wheel has 10 hollows such as 19M and the pinion T having two diametrically opposed teeth and turning 1/4 of a turn at each step, the 6A wheel will be driven every other time and will turn 1 / 10th of a turn, or 36 degrees, for each half-turn of pinion T.

   Each pinion T can therefore occupy one or the other of two positions, position I if, at the next movement of 90 degrees, it must drive the wheel 6A, and position 0 if, at the next elementary movement, it must not drive it. This pinion being connected by gears to the sun gear 2M, it follows that this sun gear will also occupy a position 0 or a position I, the position 0 corresponding to the "not retained" position! and position I to the "retained" position.



  These positions are also represented by the fingers 11M of wheels 8B cast on the planets 1M which close the contacts 30A, 30B .... when the element is in the retaining position and open it when it is in the position. - tion 0.



   The subtraction can be done with this counter following the same principle which was described previously, by inverting the position of all the transfer gears T, by adding the number to be subtracted and by inverting a second time the position of all the elements T. These reversals amount to putting in position 0 all the elements T which are in position I and in position I all those which are in position 0.

   In the example described, this inversion is obtained by actuating the "inversion" electromagnet which, by means of the rod 1614 which it pulls towards the right and levers 17M articulated in 18M on the frame, causes the upward movement of the axes 12A-12C and downward of the axes 12B and 12D, which are the axes carrying the transfer gears T, free in rotation but kept in translation. more than the normal level Tb have a lower level Tc comprising two diametrically opposed teeth 23M and 29M, like 27M and 28M of the level Tb, but offset by 90 degrees with respect to that of Tb.

   The operation of this electro will therefore ultimately have the effect of bringing the Tc levels of the pinions to the position of engagement with the wheels 6A and, due to the offset of the teeth of these pinions, the stages which were in position 0 will behave. as if they were in position I and vice versa as long as the reversing solenoid remains actuated. The return to the normal position is ensured by the springs 13M.



   For the installation at the start of the meter, in order to change the justification, it is necessary to change the position of certain stages of the meter without reacting on the following stages. Figure 25 shows how this result can be obtained. The counter shown has five stages comprising ratchet wheels 5A ... 5E by which the inputs are made and which are actuated, the electrodes 24A..24E for the inputs ranging from 1 to 16 units. The fingers 11-1 .... 11-16 control the positions of the "back and forth" 30-1, 30-2, .... 30-16. A series of 25A switches ....



  25E allows you to choose the stages which will be activated. The positioning is obtained by activating the electro of "positioning !!, which causes the displacement to the left of the rod 161,1, the descent of the axes 12A-12C and the rise of the axes 12B- 12E. In this way, the transfer gears T move relative to the wheels 6A while keeping the latter locked and the levels Ta without teeth replace the levels Tb to come into contact with the wheels 6A and deductions will no longer be transmitted.



  The toothed gears 9A-9E have sufficient width to remain in constant mesh with the wheels 8A ... 8E. The switches 25A ... 25E can occupy one or the other of two positions, the high position corresponding in the figure to position 0. It has been assumed in the figure that all these switches are in position 0, that is to say that 1-'we want to put the counter to 0. In figure 25, the counter marks 25, stages 1, 8 and 16 being in position I as indicated by the position of contacts 30A, 30D and 30E.

   The meter is put in place by activating the installation electro, by lowering it. key K which applies a battery to the electros 24A, 24D and 24E, raising the key K, which causes the advance of one step of these electro-

  <Desc / Clms Page number 24>

 tros who reset their floors to 0, and finally by releasing the electro setting up.



   The mechanical meter which has just been described being provided with the same contacts as the electric meter described above can replace the latter without modifying the circuit and the description of the operation of the overall circuit will not be repeated.



   It is clear that this counter can also be used for other purposes in different applications.



   Light decoder - Figures 26, 27, 28 and 29 show a variant of the photographic device. According to this variant, the slit ensuring the photograph of the character at the precise instant at which it passes into the shooting position by means of a photoelectric cell, is combined with the decoder. For this purpose, there is provided for each character not a single slot cooperating with a decoder, but a system of several opaque lines arranged according to a code. An optical device projects instead of a single light slit, a combination of light slits the reverse of that. opaque lines corresponding to the code of the character to be photographed. When the combinations are identical, no light passes through and the photoelectric cell gives a pulse triggering the discharge tube.

   Figure 26 shows the arrangement of the slots and corresponding characters on a portion of the disc (or drum). The background is transparent and each character is associated with a series of opaque M lines. The characters are preferably transparent on a black background (see Figure 29). The opaque lines M are arranged in seven rows T1 to T7 in a manner similar to that shown in figure 10 for the insulating and conductive parts of the decoder. Cooperating with these opaque lines is a light slit SL which , during the rotation of the disc (to the right in the case of the figure), sweeps the opaque lines. This light gap can be partially blocked by seven shutters S1 ... S7. The corresponding mechanical arrangement is shown in figure 27.

   L is a filamentary filament lamp, Cl a cylindrical capacitor making an image of the filament on an elongated window or FY slit. This normally illuminated slot can be partially closed by shutters Sl, S2, S3, S4, S5, S6, S7 controlled by solenoids or electromagnets MS1 ... MS7 against the action of the return spring Sp. An objective C2 makes an image of the slit FY on the disk D at the level of the opaque lines M. PC is a photoelectric cell and FL the light discharge tube. Suppose for example that the shutters S2, S4 and S7 are pulled by the corresponding appliances.

   It can be seen (figure 26) that, when the disc rotates, the first position for which there is complete blockage of light, is the character "x" which has the opposite combination, and prevents light from passing to the Tl levels. , T3, T5, T6.



  The optical system projecting light on these same levels, there is complete blockage. Other blockages may occur for characters following the "x", but as explained previously, they will have no effect, because the flashlamp circuit can only give one discharge per passage of the alphabet. FIG. 28 shows the overall arrangement of the organs. In addition to the components already mentioned and which are designated by the same indices, this figure shows the C3 shooting objective making an image of the character on the sensitive film SF carried by the carriage FC, the advance of which is controlled by the variable advance mechanism EV under the control of the width translator TL and a Cal cam. The ENR recorder controls the MS1-MS7 shutters through the Ca2 cam.

   The control typewriter and its selection bars are represented in CL and CM and the supply circuit of the discharge tube in PH. In addition, as shown in FIG. 29, the disc carries two alphabets of different fonts F1 and F2 each occupying one half of the periphery of the disc. The passage from one to the other is made by means of the electromagnet FSE which closes one or the other of two transparent semi-circles FIT or EST cooperating with the luminous slit SL.



  It is clear that when the EST semi-circle is closed, the photocell will receive light for the duration of the passage of the characters.

  <Desc / Clms Page number 25>

 res F2 corresponding to the semicircle FIT and, therefore, no flash of shooting can occur for any of the corresponding characters. On the contrary, the characters Fl corresponding to the semi-circle F2T can be projected. Switching from one cast to another is controlled by a special signal from the recorder which acts on the FS cast change electro.



   The opaque lines can be carried on a disc or drum independent of the die holder, but rigidly connected to it.



   First variant of justification - Figures 30 and 31 represent a first variant of the justification device.



   In this embodiment, the counter controls the operation of relay RC1-RC5 by the CO wires connected to the various stages. When the operator lowers the FLC key (figure 30) the step switch goes to position 1.



  In this position the RC1-RC5 relays which receive a battery by the counter stages which are in position "0" are blocked on FLK and release the counter. The position of the relays represents in binary numbering the deficit of the line (supposed to simplify the figure not to exceed 31, but that one can bring to any desired value by increasing the number of stages), and they serve to order the placement of a 32-position DS stepper, each level of which comprises two crowns of spindles as is well known.

   The first relay RC1 selects one of the brushes of the positioning crown, the other relays select one of the 16 pins in each half-crown o Depending on the position of the RC1-RC5 relays, one of the stepping pins DS receives land whose position precisely matches the value of the deficit. The DS stepper is activated by the battery supplied by FLK to its winding, through the resistor rs, its rest contact, and the slow release relay RA. When the Bal or Ba2 brush meets the pin connected to the earth by the relays, the stepping stops because its winding is then connected to the earth on both sides. For the sake of example, it will be assumed that the line deficit is 19 units. The broom Ba2 will therefore stop on pin 19.

   During the DS stepping operation, FLK has been kept actuated by the working contact of the slow release relay RS through which the DS working current passes, When DS stops, FLIC drops out and passes in position 2. Simultaneously FLK in position 1 has applied a battery to the brushes Bbl, Bb2 Bb3 and Bb4 of the step by step interval counter CI. This is in the position corresponding to the number of intervals which will be assumed to be 8. It can be seen that this battery is transmitted to pins 7, 15 and 23 of the row corresponding to the Ba3 brush of DS. No other pin receives potential.

   It can be checked that depending on the position of CI a battery is applied to pins N-1, 2N-1,
3N-1, etc. .. levels corresponding to the brushes Ba3, Ba4 or Ba5, N being the number of intervals. When under the control of the RC1-RC5 relays the DS brushes moved from the zero position to the 19 position, the Ba3 brush passed over pins 7 and 15 which were connected to the battery, and transmitted it. . to the step by step SZ which has thus advanced two steps, and is thus separating the piston of the variable advance 6, that is to say two units more than the normal space of 4 -units. FLK show 2, RC1-RC5 relays drop out. As soon as these relays were in operation, the counter was released for entering the next line.

   The justification is done in a manner analogous to that explained previously. A battery is for each interval applied to the juice wire by the T1-T7 translator relays (see figure 4 to 9) and a variable exhaust piston selected by the SZ switch (pist wires) at the same time the stepping DS takes a step forward each time. At the end of the 4th interval pulse, DS goes to position 23 in which his brush finds a battery applied by CI, SZ is attracted. SZ thus remains attracted until DS has fallen back to position 24, at the end of the 6th pulse.



   At this moment SZ goes to position 3 where it selects piston 7, ie the normal interval plus three units. The first five pulses are found to be 2 units and the last three to 3, giving a total of 19.



   The end of line signal arriving through the fl wire causes FLK to pass

  <Desc / Clms Page number 26>

 in position 3. In this position, FLK connected to the source of pulses G and completes its rotation. During this return to zero it fulfills the functions de- veloped at the FLR switch in the description of figures 4 to 9 such as advance of the film, maintaining the relay SE, and further returns the switches DS and CI to their rest position, by their well-known doorbell return circuits.



   Second variant of justification - FIG. 32 represents a second variant of the justification device. In this example, the step-by-step ES is set up by the relays RCl-RC5 controlled by the wires Co coming from the counter as in the case of the previous example. When the operator actuates the FLC key, the FLK combiner goes to position 1 where it supplies the blocking battery for the RCl-RC5 relays, freeing the counter for typing the next line. Suppose that, in this case again, the line deficit is 19 units. The step-by-step ES stops in position 19. At each of its positions this switch introduces two resistors in a circuit going from the battery to the earth and constituting a potentiometer by which a potential is applied to the wire k.

   This potential is chosen to be proportional to the number of units in the line deficit (plus one), for example 1 volt per unit. In the example chosen, the point k would therefore be brought to a potential of: 19 + 1, ie 20 volts.



   Moreover, the interval counter CI is in the position corresponding to the number of intervals in the line, for example 8. By means of resistors connected to its various pads, this switch sends in a set of resistors Rf , a current such that in each resistor Rf the potential drop is precisely equal to N volts, N being the number of intervals. In the example chosen, this potential drop will be 8 volts in each of the resistors Rf.



   As soon as ES has reached its position, 19 in the numerical example chosen, FLK, which was kept actuated by the work contact of the slow release relay RA, drops back and goes to position 2 where it sends a battery to the control circuit. step-by-step SZ bell operation through the contact of the polarized relay Rpl. Indeed, this relay, connected between point k at 20 volts and the brush Bc of RP2 to the ground, works. SZ starts up and the potential of the brush Bc passes successively to 8, 16 and 24 volts. At this moment the Rpl relay releases because the current changes direction and the Rp2 relay operates. The SZ stepper stops in its third position connected to the piston 6 giving a normal space advance plus 2 units.

   FLK which was maintained by the contact of Rpl at work falls and goes to position 3 where it remains during the printing of the line. The justifiable space signal arrives through the juice wire and is applied to the ES switch which advances one step each time. At the same time, a variable exhaust piston operates, this one being selected by the switch SZ as explained previously. The step-by-step ES thus goes to position 20, then 21, 22 and 23 and the potential of the point K increases by 1 volt with each step, it reaches 24 volts when ES goes to position 23 at the end of the fourth pulse, at this moment, the relay Rp2 drops because it is on each side at the same potential (24 volts), and it closes its rest contact.

   The fifth impulse therefore actuates both SZ and FLK through Rp2 at rest and SZ goes to position 4 where it selects an additional feed of 3 units for the following intervals. FLK goes to position 4 where it remains. The following pulses will therefore give intervals having one more unit, that is to say in total 5 times 2 and 3 times 3, or in total 19 units. When the end of line signal arrives on FLK through the wire fl, FLK goes to position 5 where it is taken by the pulse source G and ends its rotation during which it fulfills the auxiliary functions of the film advance. etc. .. and returns the CI counter to zero.



   - Figures 34 and 37 represent a variant of the photographic device due to Mr. Samuel H. Caldwell. According to this variant, the photographic film, instead of being carried by a movable carriage which moves after the photograph of each character, is stationary during the printing.

  <Desc / Clms Page number 27>

 line, and the image is moved by means of an optical system placed between the die-carrier disc and the sensitive film, without its dimensions being changed and without the sharpness being altered by a end of the line.



   Referring to Figure 34, the character disc is shown at D and continuously rotates in the direction of the arrow. The L flash lamp illuminates the selected character as it passes in front of a 1-P lens of focal length f1. This lens is placed at a distance from the disc, equal to its focal length and acts like a collimator giving an image of the character to infinity. On the optical axis of this 1-P lens, there is a second 2-P lens of focal length F2 associated with a 3-P prism and mounted in a 4-P mobile support, which can be removed. - place along the optical axis of lens 1-P in tube 5-P. This second lens forms an image of the character at its focus.

   The sensitive film
FP is placed at the focus of this 2-P lens, for this, the optical distance between the FP film and the 2-P lens, taking into account the reflection in the 3-P prism, is equal to the focal length of the 2-P lens. It can be seen that, in all the positions that the lens 2-P prism 3-P assembly can take, between the extreme right-hand position shown in solid lines and the starting position shown in dotted lines, it will always be a clear image on the film, an image whose dimension remains constant and depends only on the focal lengths of the two lenses. Indeed, let's call
Dl the distance, supposedly extremely large, at which lens 1-P makes an image of the character carried by the disc.

   This image is, for the 1-P lens, an inverted real image located on the right of the figure and the magnification is equal to Dl / fl. This image is for the 2-P lens a virtual object located at a distance D2, also very large, from which it gives a real image in the same direction at its focus with a magnification equal to f2 / D2.



   When the character carried by the disc is exactly at the focus of the lens 1-P, the distances D1 and D2 become infinite and can therefore be considered as equal, the distance between the lenses becoming negligible by comparison, and therefore the resulting magnification, which is the product of the two magnifications of the two lenses is equal to
D1 X f2 = f2 that is to say constant for all the positions of the len f1 D2 thread 2-P when it slides in the tube 5P. Due to the presence of the prism, the image is turned over in relation to the assembly described above, but the characters can be given the desired position by arranging the matrix characters in an appropriate manner on the disk D.



   You can change the magnification, and hence the size of the characters, by changing the focal length of the 1-P lens or the 2-P lens. To obtain larger characters, for example, it will suffice to increase the focal length f of the 2-P lens or to decrease that of the 1-P lens. In practice, it seems easier to change the 1-P lens. The new lens will be placed so that the characters on the disc are in its focal plane. Under these conditions, the image will continue to be made on the film FP and will retain a constant dimension whatever the position given to the moving lens 2-P and to the associated 3-P prism in the tube 5- P.



   Figure 35 is a side view of the device schematically shown in Figure 34. The sensitive film FP comes from a 6-P feeder reel and arrives on the 8-P feeder-receiver which is actuated at each end. line to space the lines between them, then the film is stored on the take-up reel 9-P. This assembly, as well as the movable lens-prism system is mounted on a 10-P frame and the whole is protected from external light by a suitable device, not shown.
Figures 36 and 37 show how the system which has just been described can be mounted in place of the film carrier shown in Figure 4.

   Screw 17-F, which is controlled by the variable exhaust (fi

  <Desc / Clms Page number 28>

 gures 3 and 4) is used in the same way that it has been described in relation to Figures 3 and 4, to move a nut E. The prism 3-P is mounted in a support 11-P provided with rollers 12- P bearing on 13-P rails. Figure 36 is a vertical center section of the prism carriage.



   The optical system which has just been described has the advantage of making it possible to use a stationary sensitive surface. The moving parts are reduced to a lens and prism of comparatively extremely low weight which can be moved easily, more quickly, and over greater distances, than a cart carrying a film magazine necessarily of considerable weight. higher. It also becomes possible to provide with this device magazines that can contain large quantities of film and to provide them with accessory devices such as knives, for cutting the exposed film, punches, shutter housings, summer ...



   Although the present invention has been described in relation to exemplary embodiments, it is clear that it is in no way limited to said examples and is capable of numerous variations and modifications without going beyond its scope.


    

Claims (1)

ABSTRACT. The present invention relates to a typing machine in justified characters, in which are described the various constituent elements of such a machine: recorder, counter, decoder, writing carriage, variable exhaust system and correction system. , as well as variants of decoder, justification, photographic device, and a mechanical counting system as well as the various circuits of connection of these elements.