779,521. Automatic exchange systems. STANDARD TELEPHONES & CABLES, Ltd. (Bell Telephone Manufacturing Co., Soc. Anon). June 26, 1953, No. 16594/52. Class 40 (4). In a register-controller system certain functions normally performed by a register-controller are delegated to further common circuits termed " governor " circuits, which are retained in use only while those functions are being performed. In the exchange described there are three types of such governor circuits, of which the "A" governor serves to connect up a free register which has access via a link circuit, and second and first line finders to a calling line and to connect the calling line to the register. The "B" and "C" governors assist the register in controlling all local group selections, and the final selection, respectively. Initiation of call. Closure of the calling loop brings the c wire (Fig. 1) from - 48 v. to - 16 v., and the 100-point scanner of Fig. 1 transmits a pulse in a time position characteristic of the calling line to CT1I (Fig. 5) in the first line finder/final selector electronic selection and bar operation circuit (1st LF/FS ESBO), which repeats the pulse to the grid of VT1D (Fig. 2) in the call detector. Timing pulses from source d2 (see Fig. 63) are also applied to this grid, and when a timing pulse coincides with a calling pulse, G2D (Fig. 2) is unblocked and VT1D is driven towards cut-off, so that a positive pulse appears at its anode. FT1D and FT2D (Fig. 2) form a bi-stable triggered pair of which the latter is normally conductive. The output pulse from VT1D unblocks G28D whereupon FT1D and FT2D change-over and the potential of the cathode of CT2D rises, thus indicating a calling condition. Test for a free "A" governor. The scanner connected to point CLD (Fig. 3) continuously scans the terminals B for free "A" governors. This scanner is controlled by pulse sources Nc1-4, Ne1-3, and Nd1, Nd2, Nd10 and Nd11 as shown in Figs. 64 and 65. Each free "A" governor scanned produces a positive pulse at CLD (Fig. 3) in its characteristic time unit, but in the absence of a calling condition G20D and G33D (Fig. 2) are conductive and absorb this pulse. The existence of a calling condition, however, blocks G20D and G33D and the next pulse received is then applied to the grid of VT2D (Fig. 3). Coincidence between this pulse and a timing pulse d3 causes the normally cutoff VT2D to conduct, and CT3D repeats the pulse from its cathode to trigger the flip-flop circuit FT3D, FT4D. Rectifier G22D conducts and the anode.potential of FT2D (Fig. 2) is forced down, thus restoring FT1D, FT2D, whereupon CT2D is again cut-off, and further " A " governor pulses are absorbed. After about 300 ms. condenser CI4D (Fig. 3) has charged so that FT3D conducts and the flipflop FT3D, FT4D is restored. The triggering of this flip-flop results in the charging of C3D (Fig. 2) so that G3D and G4D are grounded at both ends and the timing pulses d2 cannot reach VT1D. This tube cannot therefore respond to another calling condition until C3D has discharged after the flip-flop has restored. The positive pulse produced at the cathode of CT2D (Fig. 2) indicative of the calling condition is fed via C17D (Fig. 3) to the junction point of rectifiers G25D, G26D and resistor R65D and is differentiated by C17D and G26D. The positive-going pulse corresponding to the leading edge of the original pulse is absorbed by G26D, but the negative-going pulse (trailing edge) blocks G26D and unblocks G25D, whereupon the flip-flop FT5D, FT6D is triggered. After about one time unit (200 Ás.) C21D has discharged sufficiently to permit FT6D to conduct again, and the positive pulse produced at its anode is repeated by cathode follower CT4D to unblock G34D. The next d3 timing pulse thus passes to the anode of FT5D and the grid of FT6D to speed up the restoration of this flopflop. The resulting positive pulse at the cathode of CT4D, which is exactly one time unit long, is transmitted over terminal α (Fig. 3) to the free " A " governor scanned. When FT3D, FT4D restore after about 300 ms., the seized " A governor (Figs. 8-22) will have connected ground to the junction of G3D and C3D (Fig. 2), as described below, thus preventing the discharge of C3D, and hence the use of the call detector for another call, until the calling subscriber has been switched through. Seizure of " A " governor. This is controlled by the ambassador circuit AMA (Fig. 11) of the " A " governor which signals the condition of the " A " governor to the call detectors, receives the seizure pulses, and removes the free condition potential of the "A" governor when a seizure pulse has been received. Each circuit AMA is associated with a second scanner having a branch connected to the α lead (Fig. 3) of each call detector. As long as the "A" governor is free the two triodes CT5, CT6 (Fig. 11) are conducting and their cathode potential (- 12.5 v.) is applied to terminals # (Fig. 3) of the call detectors. The grid of VT1 (Fig. 11) is clamped at - 40 v. by sources Nc1, Nc3 and battery over R279A (Fig. 10) and rectifier G32 (Fig. 11), but a seizure pulse appearing at m blocks G32. The resulting positive pulse at the cathode of VT1 is applied to the grid of CT7 which conducts, thus reducing the potential on the grids of CT5, CT6. The resulting rise in the cathode potential of CT5 and CT6 is applied to the cathode of CT7 so that CT5 and CT6 are quickly cut-off, thus removing the free potential from terminal p. Condenser C9 charges, raising the grid and anode potentials of CT7 until the anode current drops, the resulting rise in anode potential rendering CT5 and CT6 again conductive and restoring the AMA circuit. This would normally occur within a few time units, but the positive pulse at the cathode of VT1 results in the firing of tube BAVA (Fig. 14) which lowers the potential at terminal o (Figs. 10, 11), thus unblocking G5, so that the potential of the grids of CT5 and CT6 cannot be raised sufficiently for these triodes to conduct fully. The busy condition is further maintained by the release of SfrA (Fig. 8) which occurs prior to the operation of BmrA (Fig. 10) which disconnects the potential from terminal o. SfrA in releasing disconnects the H.T. supply from CT5 and CT6 and prevents them from conducting. Continuing with the seizure, with AmrA (Fig. 18) normally up and ZurA (not shown) and BurA (Fig. 15) down, SfrA (Fig. 8) is normally operated, so that, with VorA (Fig. 14) down, the output pulse from terminal q (Figs. 11, 10) is fed to the comparator regenerator GRG1A (Figs. 12 and 13). The pulses supplied by source d3 at the beginning of each time unit to the primary winding of CALA (Fig. 20) are applied via terminal w1A (Fig. 12) and C16 (Fig. 13) to the grid of CT1. With relay ParA (Fig. 14) unoperated, CT1 conducts and the resulting pulse from its cathode is applied to the grid of VT2. If a pulse is simultaneously received on v1 (Fig. 13) from the calling call detector, VT2 conducts so that the flip-flop FT1, FT2 is triggered (FT2 then remaining conductive while C21 is charged) and is quickly restored by the next pulse from d2 applied to the grid of FT1. As d3 pulses occur at the beginning of a time unit the regenerated pulse at the anode of FT2 occurs at the beginning of the time unit following that in which the calling pulse occurs. This pulse is repeated by cathode follower CT2 to the recorder of Fig. 14. In this recorder one tube in each of the groups VbA1-4, VcA1-4 and VeA1-3 is fired indicative of the call detector calling, and operation of their anode relays brings up corresponding relays of RdarA-RddrA, RcarA-RadrA and RearARecrA (Fig. 15), under control of which the connecting relay RA (Fig. 8) relevant to the call detector is operated. The firing of BAVA (Fig. 14) to the regenerated pulse from CT2 brings up XrA, and the resulting drop in potential on lead y1 (Fig. 13) renders G30 conductive to prevent the d3 timing pulses from reaching CT1, which is thus disabled as long as BAVA remains fired. VorA (Fig. 14) pulls up, disconnects the AMA circuit (Fig. 10) from the regenerator CRG1A (Fig. 12) and brings up BurA (Fig. 15) which releases SfrA (Fig. 8) to maintain the "A" governor busy. Test of 1st LF/FS ESBO. The free condition is indicated by + 24 v. extended over a front contact of the normally operated guard relay CfrI (Fig. 7) to the sequence test circuit STA (Fig. 18) in the "A" governor. This potential is applied to the junction of G51A and R169A (Fig. 18) forming part of a gating circuit including G52A, G53A which are controlled by pulse sources Ra, Rb, respectively, individually allocated to the "A" governor. In the absence of a pulse, G52A and G53A have their cathodes at ground potential and hence the grid of OT1 (Fig. 19) cannot rise above ground potential. However, in the time position allocated to the " A " governor, they are both blocked and C41 is charged to raise the grid of OT1 to about + 16 v. Timing pulses d3 are also applied to this grid at the beginning of each time unit and if C41 has been charged in the preceding time unit, OT1 conducts. The resulting potential drop at its anode induces in the secondary of TR a voltage which causes the normally cut-off OT2 to conduct. The further anode drop resulting produces a positive pulse of about 60 v. at the grid of OT1 which fires SVA (Fig. 18), whereupon the current flow through G51A removes the free condition potential in the ESBO (Fig. 7). Relay TrA (Fig. 18) pulls up and is followed by BmrA (Fig. 10) which extinguishes the fired recorder tubes and releases their anode relays. XrA (Fig. 14) is released and BAVA is extinguished. ZxrA (Fig. 15) operates (short-circuit removed), holds in series with the operated relays RdarA ... , RcarA ... , and RearA ... , and brings up CprA (Fig. 10) which switches the "A" governor through to the ESBO. SVA (Fig. 18) is extinguished, TrA, BmrA (Fig. 10) and VorA (Fig.