中科微点-全球专利检索

  1. Login
  2. Sign Up

Search Results

4 records (took 0.02 seconds)

    1.
    发明专利
    未知

    公开(公告)号：DE1003797B

    公开(公告)日：1957-03-07

    申请号：DER0013871

    申请日：1954-03-26

    Applicant: RCA CORP

    Inventor： STUART-WILLIAMS RAYMOND ,  RAJCHMAN JAN ALEKSANDER ,  ROSENBERG MILTON

    IPC: G11C7/02 ,  G11C11/06

    Abstract: 763,100. Electric digital-data-storage apparatus. RADIO CORPORATION OF AMERICA. March 18, 1954 [March 26, 1953; May 8, 1953], No. 7940/54. Class 106 (1). [Also in Group XXXIX] In a magnetic matrix memory arrangement comprising magnetic cores, driving coils coupled to the cores to drive them selectively into .one or the opposite polarity of magnetization, and a reading coil inductively coupled to all the cores, electric integrating means receive and integrate the output signal from the reading coil obtained when magnetizing a given core for detecting its magnetic condition. Each of the cores 110 of the matrix shown in Fig. 6 is coupled to two driving coils, a row coil 112 and column coil 114, and a common reading coil 120. The coils 112, 114 are each connected to a further core 116 or 118 of a row-or column-selecting matrix switch. These cores are coupled to row and column coils 112x, 112y and 114x, 114y connected as anode loads for tubes 132x, 132y and 134x, 134y. To select a core 110, a pair of cores 116, 118 are selected by applying positive priming signals to the control grids of selected tubes. The selected tubes are simultaneously rendered conductive by a signal from source 140 applied to second control grids and drive the corresponding cores 116, 118, 110 from an " N " to a " P " saturation condition. When the signal from 140 ceases, the cores 116, 118 are restored to condition N by coils 112x, 122y connected to bias batteries, and if this restoration occurs simultaneously the selected core 110 also is driven to condition N. If it is desired to retain the core 110 in condition P, the y-switch tubes 132y, 134y are inhibited by a pulse from source 142, while the selected x-switch tubes are still conducting, so that the cores 116, 118 are restored sequentially. The sense of the windings of coil 120 on successive cores 110 is opposite so as partially to cancel out the signals from half-excited cores in the selected row and column. Curves 150 and 152, Figs. 9A and 9B, indicate respectively the contribution from non-selected, half-excited cores and the total signal induced in a reading coil when a row and column of cores are driven through a full hysteresis cycle. If the selected core was originally in condition N, Fig. 9A, equal and opposite signals are produced and the final integrated signal (indicated by broken line) is zero. If the selected core was in condition P, Fig. 9B, there is a substantial contribution from this core only during the second half cycle so that a significant integrated signal is obtained which, as well as providing an output, may be used to control restoration of the selected core to P. Fig. 10 shows a matrix system and associated reading and writing circuits employing a 5-step cycle of operations, Fig. 11. The matrix 160 is controlled by row and column switches 162, 164 driven by tubes 166x, 168x and 166y, 168y, and the cycle is controlled by the outputs of four univibrators 172, 174, 176, 178 which are operative during different steps as indicated in Fig. 11. Univibrators 172, 174 are operated directly from a pulse source 170, and 176, 178 are operated through a differentiating circuit 175 when 174 returns to its quiescent state at the .end of step 2. For reading a selected core, a read pulse embracing steps 1-4 is applied from source 188 to an " and " gate 190, thus enabling the output of 174 to open gate 192 during steps 1 and 2, and to a clamping circuit 198 (described below) to unclamp the integrating circuit 196. The output of 172 is applied directly to tubes 166x, 168x, and through combiner or " or " gate 180 to tubes 166y, 168y so as to drive the selected core to P during step 1, the core being driven to N by the bias of the switches during the following step 2. The signal in reading coil 161 is passed through amplifier 194 and gate 192 to the integrating circuit 196. If the selected core was originally in condition P, an integrated signal, of sufficient magnitude (positive or negative according to the sense of the reading winding on the core) to pass upper or lower discriminator 200 or 202, is present at the end of step 2 (see Figs. 9B and 11). Such a signal is applied through " and '' gate 204, opened by univibrator 178 during steps 3 and 4, " or " gate 186 and " and " gate 182, also opened by 178, to gate 180 and tubes 166y, 168y to hold the selected core of y switch 164 in condition P until the end of step 4. This ensures that the x and y switch cores are restored sequentially, at the beginning of steps.4 and 5 respectively, after the selected matrix core is again driven to P during step 3 by the output of 176 (in place of 172, step 1). For writing condition P into a selected core, the signal from gate 204 is replaced by a write pulse applied from source 185 through " and " gate 184. In the absence of the write pulse, the switch cores are restored to N simultaneously during step 4, and N is written into the .selected matrix core. In a modification, writing P is effected by shortening, instead of lengthening, the y switch pulse, inhibit gates being employed. Reference is made to a three-dimensional arrangement. Integrating and associated circuits. The circuit 196, Fig. 10, comprises a tube 220, Fig. 13, and a feedback circuit including condenser 222 and cathode follower 224. The anode potential of the tube is normally clamped by a circuit (198, Fig. 10) including diodes 230, 232, 234, 236 and a pair of triodes 238, 240 having a common cathode resistor 254, the triodes being normally so biased that 238 is cut off and 240 conductive thus enabling the diodes to conduct and reproduce the required clamping potential, applied at junction 248, at junction 246. A negative pulse applied to the grid of 240 in response to the read pulse from 188, Fig. 10, produces an exchange of conduction between 238, 240, thus blocking the diodes and freeing the integrator tube 220. The discriminators 200, 202, Fig. 10, comprise pairs of triodes 260, 262 and 256, 258 respectively, Fig. 13. The grids of 258, 262 are so biased in relation to the normal integrator output potential at the cathode of 224, that 256 and 262 are normally conducting. If the integrator output falls below or rises above the predetermined bias voltages (50 V. and 190 V.), an exchange of conduction occurs, and triode 258 or 260 conducts to produce a negative output across the common anode load resistors 268. In another matrix system (Fig. 9, not shown) the output signal from the reading coil is applied to the integrating circuit through a non-linear transformer, which discriminates against smallamplitude signals, a push-pull amplifier, and a rectifying and clipping circuit; also, the integrator tube is gated through a delay network so as to commence integration a short interval after the application of a driving pulse to the matrix.

Patent Agency Ranking