`
`
`
`
`UNIFIED PATENTS
`
`EXHIBIT 1007
`
`
`UNIFIED PATENTS
`
`EXHIBIT 1007
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 1
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 1
`
`

`

`ISept. 23, 1952
`
`T. K. SHARPLESS Er AL
`MAGNETIC DATA STORAGE SYSTEM
`
`2,611,813
`
`Filed May 26, 1948
`
`7 Sheets—Sheet 1‘
`
`Fm.
`
`570R66§
`unvr
`
`
`
`
`JTMHGE fo/PflE/YT
`T/Ffi/I5/‘7/55/0/7 fey/PHE/VT
`
`
`
`
`throw/1
`W
`
`
`57,4770” EQUIP/75777
`
`
`
`
`Tfifi/YJN/Sf/O/Y [Ml/ONE”7'
`
`
`
`
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 2
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 2
`
`

`

`Sept. 23, 1952
`
`T.’ K. SHARPLESS EEAL
`MAGNETIC DATA STORAGE SYSTEM
`
`2,611,813
`
`Filed May 26, 1948
`
`‘7 Sheets—Sheet 2
`
`l
`
`/qu(/%x5
`I ”\
`:llflll
`FranAmi
`Ifir‘.
`lab In“ I1I Db/flZa/orfi .9.
`
`”illtli [-
`
`+V_ F767. V_
`
`'
`
`Vr
`
`7/”,sz '
`
`+,V._flV_ mm?
`E i
`I
`4150
`
`‘9
`{.é‘I-I-I1_____ F.
`"I35 Ilia! I'é
`'% K
`3",
`II;;:
`a
`
`
`A
`
`I
`
`‘
`
`.MTT
`O) J I
`
`3:
`
`VopVVsé/mu
`Fm a
`
`+2
`+2
`
`
`
`6: ”if
`
`-
`
`'
`+2 +2'
`-- +75
`+2 +3.
`.25ipzvi257:—-
`mwfmk("V/'7‘” J'Wjfifim Z’exa'
`122423215019ch
`Eg/iflkzrléfififi?"
`.5
`W/fiwwn/
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 3
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 3
`
`

`

`Sept. 23, 1952
`
`.
`
`T. K. SHARPLESS ET AL
`MAGNETIC DATA STORAGE SYSTEM
`
`2,611,313
`
`Filed May 26, 1948
`
`7 Sheets-Sheet 3
`
`+2/0
`
`f/50 +2/0
`
`Hg. 5.
`mwr
`
`Xmm
`
`Ef’
`
`loo/r
`
`[rose 5100/30/(6'.
`)(mw draw/4752'(.944
`+/50-
`
`?
`
`390K
`
`\—“ __—?_—
`75 Q
`
`I
`
`_____7b—Pair—r;a_/
`5
`
`fima’
`
`
`
`V?! ’00”
`
`qfi. 6:
`/50PqueJ/rex_mm oé
`
`
`‘
`fey/Lifer
`L” 2:0”
`hi
`
`
`
`
`
`J
`
`7’ .
`
`flaw-1mm”
`IWWJTIMZ‘J
`gyfieflfllfi? 5
`[5. me1%an fibhwgwo
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 4
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 4
`
`

`

`Sept. 23, 1952
`
`T. K. SHARPLESS ET AL»
`MAGNETIC DATA STORAGE SYSTEM
`
`2,611,813
`
`Filed May 26, 1948
`
`.
`
`7 Sheets-Sheet 4
`
` 7bé‘Idre
`
`
`
`
`M11245?[Ml—J
`‘ 52%ZZ’erzgj
`&-
`W
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 5
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 5
`
`

`

`SePt- 23, 1952
`
`T. K. SHARPLESS ET AL
`MAGNETIC DATA STORAGE SYSTEM
`
`2,611,813
`
`Filed May 26, 1948
`
`-
`
`7 Sheets-Sheet 5
`
`F269
`
`+150 . MK '
`
`CYCLE}? 9
`
`I AmeJe/ec r ,‘ [4"
`Em
`9
`
`
`genera/w
`
`‘W’ Q/flwf 75 §
`
`Fry/n line
`«Je/ecfar— 0,, F7qjl
`
`.0/
`|
`
`+75
`
`\me Z/he 5e/ec/ar
`02, fig. //.
`
`Fexf Pub-e
`
`My D/S’fF/Za/or
`
`I
`
`I
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 6
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 6
`
`

`

`Sept 23, 1952
`
`T. K. SHARPLESS ET AL
`MAGNETIC DATA STORAGE SYSTEM
`
`2,611.813
`
`Filed May 26, 1948
`
`7 Sheéts—Sheet 6
`
`625/
`
`Lia
`
`'
`
`Lib
`
`./’7
`
`I\
`
`
`
`Alfie é/ecv‘or T fié’ //
`
`Lee M
`
`'
`
`5
`
`'
`
`/' W
`
`
`
`:fl‘
`
`j_‘ ° 6I5
`
`__
`
`75750—7 ,H 9,0
`
`
`
`
`
`7b Fbfl‘fig/z' I
`@yémra’ 2554.5;
`
`
`
`ZEIQhflfvw
`Ira/8222275:
`ffipmffézwjalé’fi
`IimJ'Z'xb/za'f JP.
`yfiw‘"A229“?
`
`W4
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 7
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 7
`
`

`

`Sept 2,3, 1952
`
`Filed May 26, 1948
`
`T.'K. SHARPLESS ET AL
`MAGNETIC DATA STGRAGE“SYSTEM
`
`2,611,813
`
` ----
`
`—-—
`---
`———
`--—
`---
`-—_
`
`7 Sheets-Sheet 7
`
`Fifi/.5
`
`M
`
`+I50
`5’5
`é 7"“
`
`2
`
`s(II? MI s
`
`’g
`i
`
`P/cfl’l/
`
`27
`
`/
`
`,
`
`__§£_
`.40-
`
`2; +a
`
`
`
`W T
`
`o emf/m.fi/é’cfmflqfll
`
`
`
`F76 /4 )5V-wc/J».
`Mr,‘53 "
`
`Fry/77 Other Par/Warn
`z}? 5fof/b0
`
`72/En
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 8
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 8
`
`

`

`Patented Sept. 23, 1952
`
`2,611,813
`
`UNITED STATES PATENT OFFICE
`
`2,611,813
`MAGNETIC DATA STORAGE SYSTEM
`
`Thomas K. Sharpless, Haverford, and Edwin S.
`Eichert, Jr., Springfield, Pa... assignors to Tech-
`nitrol Engineering Company, Inc., Philadel-
`phia, Pa., a corporation of Pennsylvania
`
`Application May 26, 1948, Serial No. 29,324
`24 Claims.
`
`(Cl. 177—353)
`
`1
`This invention relates to systems for storing
`information, especially where it
`is desired to
`transmit, receive and record the information.
`More particularly, the invention relates to sys-
`tems where the information is conveyed by means
`of groups of electrical impulses of the digital, or
`pulse and no pulse, sort. While the invention
`may be used for various purposes, by way of ex-
`ample it may be used to store information con-
`cerning reservations on public carriers such as
`airplane lines, railway lines, etc.
`The principal object of the invention is the
`provision of a system whereby persons at a plu-
`rality of remote positions may insert and with-
`draw information from a centrally located stor-
`age unit comprising a plurality of registers each
`of which will hold its information permanently
`unless changed by the insertion of new data. from
`any one of the positions. In addition, the storage
`unit may have associated with it an adding unit
`in order that numerical
`information may be
`stored and accumulated.
`A more specific object of the invention is the
`provision of such a system wherein the informa-
`tion is stored by magnetic recording of pulses on
`rotating magnetic disks having register areas
`which are selected through the agency of reg-
`ister-selection voltage combinations representa-
`tive respectively of the registers and occurring
`successively in timed relation with the rotation
`of the disks, the selection of the registers being
`effected through coincidence of said register-
`selection voltage combinations and voltage pat—
`terns produced by action on the part of an
`operator.
`Other objects and features of the invention will
`be apparent from the following description.
`In the accompanying drawings:
`Fig. 1 is a block diagram of an information
`storage system according to the invention;
`Fig. 2 is a perspective view showing the pre-
`ferred form of the central storage unit;
`Figs. 3a and 3b are face and side views respec-
`tively of a register disk employing a plurality of
`recording heads, this being a possible alternative
`arrangement, as hereinafter described;
`Fig. 4 is a generalized illustration of the regis-
`ter selector;
`Fig. 5 is a diagrammatic illustration of one of
`the units of the register selector:
`Fig. 6 is a generalized illustration showing the
`electrical arrangement of the adder in associa-
`tion with the register selector;
`Fig. 7 is a diagrammatic illustration of the co-
`incidence circuit d employed in the register
`selector;
`Fig. 8 is a diagrammatic illustration of the
`electronic selecting switch f of Fig. 4;
`Fig. 9 is a diagrammatic illustration of the
`cycler and the decoding distributor;
`
`10
`
`l5
`
`[3GI
`
`30
`
`35
`
`40
`
`50
`
`55
`
`60
`
`2
`Fig. 10 is a diagrammatic illustration of an
`electronic switch which is employed at various
`places in the system;
`Fig. 11 is a diagrammatic illustration of the
`line selector;
`Fig. 12 is a diagrammatic illustration of the
`encoding distributor;
`Fig. '13 is a diagrammatic illustration of the
`equipment at one position of a station;
`Fig. 14 is a diagrammatic illustration of the
`position selector; and
`Fig. 15 illustrates a possible arrangement which
`may be used in the push-button keyboards.
`
`-
`
`General description of system
`
`Figure 1 shows a generalized view of the sub-
`ject system.
`It shows the central storage equip-
`ment, the transmission equipment, and the trans-
`lating equipment.
`It should be pointed out that
`the transmission equipment consists of a plu-
`rality of lines feeding from a like number of sta-
`tions.
`In addition each station permits the use
`of a plurality of positions. Each position of the
`translating equipment has thereat a keyboard for
`inserting and requesting information and an
`indicator to display information from the storage
`unit. By setting up the proper keys the operator
`may select any one of the registers in the storage
`unit. Likewise, by the setting of other keys he
`may request the information in that particular
`register or insert new information in that register.
`That only one position in a station be operative
`at any time is assured by an interlocking circuit
`working in conjunction with the position selector,
`which guarantees that each position get its turn
`on the line. The line selector of the central
`equipment works in a similar manner to prevent
`interference between lines.
`The information for the selection of the reg-
`ister and the information to be stored are trans-
`mitted in the form of time division coded groups
`of electrical impulses, as is common in Teletype
`systems.
`In this description such a pulse group
`will be called a word. The uses of the encoding
`distributor, cycler. and the decoding distributor
`for producing the words, will be made clear in
`the detailed description of the transmission equip— ,
`ment. The register selector, working in con—
`junction with the decoding distributor, splits the
`word into its two components, using one part to
`select the proper register and sending the second
`or information part either into the register or
`adder as needed, or if it is merely a request, allow-
`ing the register contents to go into the trans—
`mission equipment.
`The drawings show, diagrammatically and
`symbolically,
`the essential components of
`the
`system as generalized in Figure 1 and as it has
`been constructed and successfully operated.
`In
`the subsequent description, the component cle—
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 9
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 9
`
`

`

`3
`vices and their functioning will be described in
`succession, commencing with the storage unit
`and proceeding through to one of the station
`positions, and then the complete operation of the
`system will be described.
`
`2,611,813
`
`10
`
`15
`
`20
`
`30
`
`35
`
`40
`
`Detailed description of the storage unit
`The central storage equipment consists of a
`storage unit making use of the principles of mag-
`netic recording for the purpose of storing the im-
`pulses. These impulses are recorded as discrete
`areas of magnetization around the periphery of
`a. thin circular disk of suitable magnetic mate-
`rial as shown in Figure 2. A plurality of such
`disks are mounted on a shaft driven continu-
`ously by a suitable motor. Referring to Figure
`2, disk I
`is used as a master pulse source, or
`clock.
`It has permanently recorded on it two
`channels, :5 and :11. Channel :1; carries a given
`number (e. g. 160) of pulses recorded at suitable
`intervals, leaving a small sector blank. Channel
`11 carries one pulse located in about the center
`of the segment delineated by the blank sector
`of channel :r. The recording heads A and B are
`located so that their air gaps cover respectively -
`channels a: and 11. Each recording head is con-
`nected to a suitable vacuum tube amplifier to
`bring the pulses to a suitable voltage level for
`operating the rest of the system
`Still referring to Figure 2, disks 2 through 71
`are used as register disks, each with a single
`head serving the purpose of recording pulses on
`the disk, reading the pulses on the disk and
`erasing pulses from the disk. The number of
`disks actually employed will depend upon the
`requirements in any given instance. Each reg-
`ister disk contains a plurality of registers chosen
`by counting the impulses from the clock disk.
`For example, in the presently disclosed system,
`each register disk contains 16 registers of 10
`pulse spaces each. These are delineated by not-
`ing every 10th pulse from the 160-pulses recorded
`on the clock disk.
`It should be noted that several channels may
`be handled on one disk by placing the heads
`around the disk with their air gaps at different
`radial distances from the edge. Figures 3a and
`3b show such an arrangement using four heads
`I to IV and four channels. The limitations on
`the number of channels thus available are set
`by the depths of the throat of the recording head,
`and by the closeness with which the magnetic
`spots may be placed on the disk.
`Description of register selector
`
`50.
`
`C! 01
`
`60
`
`In the illustrated embodiment of the system.
`the register disks are used in pairs, one disk of
`a pair being used to record units digits of num-
`bers, and the other disk of the same pair being
`used to record tens digits of the numbers. Fig-
`ure 4 is a generalized illustration of the register
`selector, there being shown four register disks
`2 to 5, of which disks 2 and 3 constitute one pair,
`and disks 4 and 5 constitute another pair. These
`pairs of disks are selected in a manner presently
`to be described.
`The clock disk I gives 160 pulses per revolution
`through amplifier a: which pulses drive I), a scale-
`of- ten electronic counter The 16 pulse per rev-
`olution output of this counter is used to aper-
`ate a counter c consisting of four cascaded scale-
`of-two electronic counters. The outputs 0f the
`four counters give 16 unique combinations of
`positive and negative voltages per revolution,
`one for each of the impulses which enter it from
`
`4
`b. These combinations are repeated each revo-
`lution. These output voltages are fed into the
`coincidence circuit d where coincidence of the
`voltage combination from c with that of the four
`.input voltages V1 .
`.V4 produces an output
`voltage of a duration of 10 pulses, which is ap—
`plied to twoof the electronic switch and ampli-
`fier units e1 .
`.
`.
`(:4, thereby activating the asso—
`ciated recording heads which are used to “Write”
`input pulses in the register, “Read” pulses from
`thevregister, or to “Erase” the pulses already re-
`corded in that register. The scheduling of these
`operations is carried out by unit M, and is ex-
`plained below.
`It can thus be seen that any one of 16 ten—
`pulse sectors around a register disk may be
`chesen by the 16 possible on and, 011 combina-
`tions of V1 .
`.
`. V4. Inthe system dis’closedhere-
`in, numerical information is stored, using a sim-
`ple linear code wherein the digit, to be stored is
`represented by a number of pulses equal to. that
`digit, i. e., no pulses for 0, one pulse, for 1, and,
`so on to nine pulses for 9. Moreover this same
`system is designed to store numbers, up- to 99
`and uses one register disk for the units digits and
`another disk for the tens digits. Thus e1 is con-
`nected in parallel with ea. so as to handle both
`digits of the number simultaneously on. the disks
`4 and 5. The block f represents a two way elec-
`tronic switch, as hereinafter described, which
`serves as a means for selecting register disks 2
`and 3 or 4 and 5, depending on V5 being on or oil.
`The one pulse per revolution output supplied
`from the clock disk I through amplifier (11 is used
`to initially set, the counters so that the registers
`on the disks will always maintain the same re-
`lation with the pulses on the clock disk as
`checked by counters b and 0, even though the
`power be shut off and later turned. on with the
`counters coming up containing arbitrary counts.
`The circuits of the amplifiers a1 and (12 are
`conventional vacuum tube amplifier circuits.
`The circuit details of the counters b and c are
`quite well known and have been described by
`Sharpless
`(Electronics, March 1948), Blume
`(Electronics, February 1948), and many others.
`The symbol K represents a conventional high
`frequency (e. g. 30 kc.) oscillator which produces
`the erasing signal. The blocks Cl and Se repre-
`sent devices whose nature and purpose will ap-
`pear later.
`Figure 5 shows the details of the, switch and
`amplifier unit e1 which is typical of all four. units.
`In addition, Figure 5 shows how the unit M
`schedules the operations of Read, Write, and
`Erase. The block f is the same one shown in
`Figure 4. The block 9 represents 4 stages—0, 1,
`2, 3—of
`linear electronic counter. The out-
`puts of three stages are used to turn- on the
`grids of the tubes T1. T2, T3 respectively. The 0,
`stage has no output used, but is connected to the
`clear circuit, h, which sets the counter to 0 at the
`occurrence of a reset pulse. The counter _g is fed;
`from the output of d of Figure 4, thus stepping
`each time a coincidence is made in d. With no
`reset pulse present, counter 9 will step- from 0 to
`1, from 1 to 2 from 2 to 3. from 3 back to 0,and
`continue this cycle as long as impulses from d
`are present.
`Still referring to Figure 5, coincidence of a
`signal from f on the first grids of tubes T1 to ,T:
`with that of the output of the corresponding
`stage of counter y will produce a negative volt-
`age swing at the plate. of the tube in question.
`The Read circuit, which involves tube T4 and
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 10
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 10
`
`

`

`2,611,813
`
`6
`5
`transmission equipment. On examination it will
`amplifier a3, is activated by the signal from tube
`be seen that only when the voltage output pat-
`T1. Tube T4 is a double triode which is operated
`tern of V1 .
`.
`. V4 and the counter c agree in
`with a slightly positive bias on each grid and
`opposite phase, as shown, will no current be
`the plate load resistor chosen so that if either
`.
`. T14 and a signal
`triode section is conducting, the level of voltage 5 drawn through tubes T7 —.
`at the common plate connection is sufiiciently
`operate T15. For example, if the V1 .
`.
`. V4 pat-
`low so as to render any subsequent circuits in-
`tern is + —~, — +, — +, + —, the counter 0 must
`operative. Only if both triode sections are cut
`give a pattern — +, + —, + —, — +, to cut off
`off will the plate voltage rise to 75 volts and op—
`all of the tubes T7 .
`.
`. T11.
`It will be remem-
`erate the following circuit.
`It can thus be seen 10 bered from the description of the register selector
`that only for the duration of the negative ex-
`that counter c only remains in any one state for
`cursion of the plate of T1 will the negative pulses
`ten pulse times and repeats its cycle with each
`applied to the other grid of T4 come out at 2. volt-
`revolution of the clock disk I; thus the signal
`age level sufficient to operate the output circuits.
`from -.tube T15 will have a duration of ten pulse
`The negative pulses arrive at T1 from amplifier 15 times and will appear once each revolution at a
`as, which receives the pulses from recording head
`different part of the revolution for each V1 .
`.
`.
`F and disk 5.
`V4 pattern.
`The Erase circuit which operates from tube T2
`Figure 8 shows the electronic switch repre—
`accomplishes a similar object, 1. e., that of al-
`sented by block f of Figure 4, and shows how the
`lowing the high frequency erasing signal from 20 ten pulse signal from T15 is switched into one or
`oscillator K (Fig. 4)
`into the driving amplifier
`the other of the two channels leading to 61 and
`Ta and thence into the head F thereby erasing
`e2 or to e: and e4. Here again, a double triode
`pulses on disk 5 only for the duration of the sig-
`coincidence circuit similar to that of the T1 read
`nal from tube T2. Here, as long as either input
`out circuit of Figure 5 is used. Negative ill-pulse
`to the right-hand grid of T5 is positive, the grid 25 duration signals from d are applied to T16 and
`is maintained positive and the consequent diode
`T17 and, with Vs as shown, will appear only at
`action efiectively prevents any appreciable signal
`the plate of T16 at the proper voltage level
`to
`appearing at the plate of the right hand triode
`operate the T1 .
`.
`. T3 tubes of the 61 and as.
`section of double triode tube T5. When the plate
`blocks as shown in Figure 5.
`of tube T2 swings negatively, the diode action 30
`.
`.
`of the grid of tube T5 ceases and large signals
`Description of the adder
`appear at the plate.
`The adder is connected so that it receives its
`The other half of tube T5 works similarly to al—
`input from two sources as shown in Figure 6.
`low input pulses to be recorded on the disk only
`One source is the decoding distributor of the
`for the duration of the signal from tube T3. The 35 transmission system and the other is the output
`1N34 crystal diodes are used in the plate circuit
`pulses from the switch-amplifier blocks, e, of the
`to prevent
`loading down of one plate by the
`register selector. The adder’s output is connected
`plate of the other conducting half of tube T5.
`to the input pulse lines of the same blocks.
`Its
`It can thus be seen that the functions of read-
`purpose is to receive numerical information from
`ing the disk, writing on it, or erasing from it may 40 the transmission system, add this to the contents
`be accomplished. The reset pulse is used to ini-
`of a register, and to transfer the sum back to the
`tially set 9 to 0 when the power is turned on and
`same register. The details of the adder circuit
`also, as will be explained below. to make it DOS-
`are not necessary here, since the adder is es-
`sible to Skip the Write or Erase operations un-
`sentially a two decade accumulator such as de—
`der certain conditions.
`45 scribed by Burks (“Electronic Computing Cir—
`As mentioned above, Figure 5 shows only the
`cults,” Proc. I. R. E. 35 : 756, August 1947) and
`apparatus associated with register disk F. Sim-
`by Brainerd and Sharpless (“The ENIAC,” Elec-
`ilar apparatus will be provided for each of the
`trical Engineering 67 : 163, February 1948). The
`other register disks. as represented by the blocks
`block 1'. represents a coincidence or switch circuit
`ezto eithlgure 4.
`'
`50 similar to those of T1 .
`.
`. T3 in Figure 5 and
`The coincidence circuit of the block :1 of Fig-
`allows a series of pulses to cycle the adder during
`ure 4 is detailed in Figure 7. The cascaded scale-
`the Write operation. This cycling is done so
`of-two counters c are shown with each half indi-
`that the numerical contents of the adder may
`eating the polarity of its output voltage with
`be transmitted as described in the above refer-
`respect to +75, the cathode level of vacuum tubes 55 ences. The symbols 9', and 7‘2 represent diode
`T7 through TM- The voltages V1 -
`-
`- V4 appear
`buffers which prevent back coupling between the
`each on two wires. One of the wires is positive
`units and tens lines.
`and the other negative with respect to +75.
`_
`.
`TheSe voltages and those from the stages of
`Specull features of the central storage equzpment
`counter c are fed to the grids of tubes T7 .
`.
`. 60
`The foregoing sections have described the cen-
`Tu, which are coincidence tubes similar in func-
`tral storage equipment. of the system. Certain
`tion to T1 <
`~
`- T3 0f Figure 5- Only five 0f the
`outstanding features are set forth below.
`.
`tubes T7 .
`. T14 are actually shown but
`the
`(1) The use of a magnetic storage medium
`presence of the others will be understood from
`which permits compact storage of impulses, quite
`the illustration. With the pattern of voltages 65 high pulse rates far reading pulses on and off,
`shown in Figure 7, it can be seen that every tube
`and easy erasure and reuse of the same material.
`of the group T7 .
`.
`- Tu has at least one of its
`This feature also has the advantage that there
`control grids negative with respect to its cath-
`will be no loss of stored information in the event
`ode, thus permitting no flow of current through
`of power failure.
`'
`the common plate load resistor and allowing the 70
`(2) The use of the disk form for the magnetic
`material. This is compact. easily assembled, and
`plate line to rise to +150. This rise in voltage
`easily produced. Moreover. the disks can be ro-
`is delivered through the 82K and 100K step
`tated at high speeds, allowing a short recess time
`down circuit to one grid of tube T15, and is suffi-
`to any register.
`cient to turn that tube on when the other grid is
`driven positive by the “go ahead” Signal from the 75
`(3) The use of more than one recording chan-
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 11
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 11
`
`

`

`2,611,813
`_
`
`7
`nel per disk which, for instance, permitsthe use
`of one pulse space per decimal digit in a 4-chan—
`nel system.
`(4) The use of a clock disk as the master
`source of pulses, which obviates all synchronizing
`problems between disks.
`(5) The use of electronic counters and switch-
`ing circuits which operate at pulse rates up to
`hundreds of thousands per second. Such high
`speed switching and counting permits the use of
`high pulse rates from the clock disk,
`thereby»
`greatly speeding the operation of the whole sys-
`tem.
`(6) The use of binary‘or base two combina-
`tions for the selection of registers. This means
`that each scale-of-two stage added to counter c
`will double the number of registers one can select.
`General description of central transmission
`equipment
`
`10
`
`15
`
`20
`
`8.
`This requirement is easily met by the proper
`choice of the values of the capacitance and re-
`sistance in, the coupling circuit to the second grid
`of T19.
`
`Details of decoding distributor
`Figure. 9 also shows the decoding distributor.
`The tubes T20 .
`.
`. T27 are coincidence tubes, as
`mentioned previously, and for simplicity only the
`two input grids and the output plate of each tube
`are shown. As counter 171. steps onto 0, 1, 2 .
`.
`. 7,
`each one of the first gridsis driven on in turn so
`that if a pulse is on the input line from the line
`selector during a particular part of the cycle,
`that pulse will appear at the plate of the tube
`whose first grid is positive.
`In this manner the
`pulses from the cycler. which have been time
`division coded into a word in the translating
`equipment of the station and are returning to the
`central equipment, are distributed either to the
`adder or
`to the coding switches SI .
`.
`. 85,
`which produce the previously-mentioned,voltages
`V1 .
`.
`. V5. It should be noted that the pulse ar-
`rives at the second grid of Tzo .
`.
`. T27 simul—
`taneously.
`It arrives there, however, delayed by
`the time of travel from T19 out over the trans-
`mission equipment,
`through the translation
`equipment, and back over
`the transmission
`equipment. Thus the pulse repetition period of
`2) must be greater than the time of travel of the
`pulse over the route indicated.
`If this cannot be
`done without too great a sacrifice of speed, as for
`instance might be necessary for use with very long
`lines, extra stages may be added to m between 8
`and 0 to take care of. the initial delay. It will be
`noted that tubes T20, T21 and T22 are connected
`together to the adder so that pulses in the first
`three time positions of the word enter the units
`decade of the adder. T23 .
`.
`. T21 are connected
`individually to SI .
`.
`. 85 and thereby convert
`the last five pulses of the word into the voltages
`V1 .
`.
`. V5 which operate the register selector
`coincidence'circuit.
`Figure 10 shows acircuit suitable for use as the
`S blocks.
`It is a typical Eccles-Jordan trigger
`circuit which has two stable states, set and reset.
`When set the M output is positive and the L out-
`put negative. When reset M is negative and L
`positive.
`A negative pulse applied to the J input will set
`the circuit while a. similar signal on K will reset
`it. The voltages V1 .
`.
`. V5 of Figs. 7 and 8 are
`produced by the L and M outputs of .91 .
`.
`. 35
`of Fig. 9.
`
`Description of the line selector
`The line selector used in the particular system
`disclosed is shown in Figure 11. Merely by way
`of illustration, the line selector is shown as be-
`ing adapted toscan periodically three lines each
`consisting of a pair of wires balanced to ground.
`The line L111 and Lu: is the one which extends to
`the typical station equipment hereinafter de-
`scribed. As shown the line selector is in the form
`of a commutator q which comprises input and
`output arms 1' and v on a common motor-driven
`shaft, and stationary contact segments engage-
`able by the respective arms. Arm 1) makes con-
`tinuous contact with ring III, while arm 1' makes
`continuous contact with ring I I. These rings
`have external circuit connections as shown. Arm
`0 also makes contact successively with three sta-
`tionary contact segments l2,
`l3 and M which
`are insulated from one another. Arm r makes
`contact successively with stationary contact seg-
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 12
`
`(Fig. 1)
`The central transmission equipment
`comprises three major components, the cycler,
`the decoding distributor, and the line selector.
`The cycler consists of a continuous pulse genera—
`(39‘1
`tor working in conjunction with a coincidence .
`circuit which is under the control of the decod-
`ing distributor. The cycler will, when released
`by a signal from the line selector, give forth a
`continuous burst of a given number of pulses.
`These pulses cycle the decoding distributor and
`also go out over a line to a station where they
`are coded up as a group of timed pulses, or word,
`in the station equipment. They then return over
`the transmission line and are distributed into
`the proper channels of, the central equipment by
`the decoding distributor. The line selector scans
`in turn each of the plurality of lines leading into
`the central equipment. When one of these lines
`is activated from a given station, the line selector
`will “lock-up” on that line and will release the
`cycler.
`It will remain on that line until it re-
`ceives a “reset signal” from the register selector
`when it resumes its scanning.
`
`30
`
`40
`
`Details of the cycler
`Figure 9 shows in detail how the cycler op-
`erates in conjunction with the nine stage linear
`counter m of the decoding distributor to give out
`a group of 9 pulses to the transmission equip-
`ment.
`The block 1: represents a conventional ,
`pulse generator which may be of the multivibra-
`tor sort, and which continuously supplies posi—
`tive impulses to one grid of tubes T18 and T19.
`Tubes T13 and T19 are coincidence tubes,
`like
`T1 .
`.
`. T3 in Figure 5. For the present purpose
`it suffices to show only the cathode, two control
`grids, and the output plate of each tube. The
`second grid of T11: is directly connected _to the
`negative output of stage 8 of counter m, so that
`it. is cut off when m is on that stage. The second
`grid of tube T19 is cut off by the bias voltage sup-
`plied through the 100K resistor. Consequently
`no pulses from p can enter m nor the output
`circuit, except when a positive impulse from the
`line selector is applied to the second grid of T19
`through the 0.01 capacitor, causing that tube to
`permit one pulse to pass. This pulse goes out the
`output and also steps the counter 71:. from stage 8
`to stage 0. Pulses now can pass through tube
`Tia until m arrives back on stage 8, when Tia will
`again be cut off and the device will
`lock up,
`awaiting another positive impulse on the second
`grid of T19. It should be noted that this positive
`triggering pulse must have the duration of at
`least one pulse time and not more than eight.
`
`00
`
`85
`
`70
`
`75
`
`UNIFIED PATENTS EXHIBIT 1007
`PAGE 12
`
`

`

`2,611,813
`
`9
`l6 and [1 with which there are asso-
`ments l5,
`ciated auxiliary segments [8,
`i9 and 20 respec-
`tively. The latter are connectable to their re-
`spective associated main segments through con-
`tacts 2|, 22 and 23 of relay 81. The purpose of
`the auxiliary segments will be explained pres-
`ently. As the arms 7 and v rotate, they “scan”
`the three lines which are connected to the sta—
`tionary segments. Each line consists of an in-
`put wire and an output wire as shown.
`The input arm r is connected (through ring 11)
`to the grid of vacuum tube Tao, which is used to
`control the relay 81. Arm r is also connected
`through to the input of tubes T20 .
`.
`. VTz'r of Fig-
`ure 9. The output arm 22 is connected (through
`ring 10)
`to the output from'Tia and T19 of the
`cycler, Figure 9. Both arms are also connected
`to the indicator circuits, T31 and T32. The pur-
`pose of these tubes is to send positive pulses out
`on the line to operate the indicator lights at the
`keyboard positions, as will be described later.
`The operation of the line selector is as follows.
`With the commutator revolving, the input arm r
`contacts a line with a positive potential on it,
`which indicates that the line is activated. This
`positive potential is thus applied to the grid of
`tube T30,
`turning it on and drawing current
`through relay 31, which picks up, opening the
`drive motor circuit and closing the contacts to
`the auxiliary segments on the input ring of the
`commutator. These auxiliary segments are vital
`to the action of this device. The commutator
`and motor have rotational inertia and will con-
`tinue to move a short distance after the power
`is removed.
`If the auxiliary segments are not
`present, the selector might pick up near the end
`of one segment and the arm coast beyond, thus
`dropping the selector again without locking up.
`However, with auxiliary segments which are
`slightly longer than the maximum coast this ob-
`jection is overcome. For now 31 will only pick
`up while the arm is on the main segment, and
`with the auxiliary segment switched in by the
`action of 31 the arm cannot coast by and drop Si
`again. Of course the contact the arm makes in
`going from the main segment to the auxiliary
`segment must be of the shorting type. As Si
`picks up, it also applies a. positive potential to
`the capacitor input to tube T19 of Figure 9, thus
`releasing the cycler as described above. The line
`selector is released by either one of the indica-
`tion signals resetting the keyboard at the station
`end of the line which removes the positive po-
`tential from the line and the grid of tube Tao,
`thus releasing relay Si. The commutator is now
`free to continue its scan.
`
`10
`
`30
`
`40
`
`45
`
`50
`
`Description of the indicator signals
`The indicator signals in the system described
`herein are quite simple. Theirfunction is to in-
`dicate whether or not any total produced by add-
`ing a number from the station to the contents
`of a register exceeds 99. The signal which indi-
`cates the exceeding of 99 is called the alarm. The
`pulse indic