US2965883A - Electronic gang switches - Google Patents

Description

ELECTRONIC GANG SWITCHES Filed Dec. 20, 1954 ,Pam/aml nfs/voeu. 5. A44/1.1.5,?,

United States Patent Oce 2,965,883 Patented Dec. 2o, 1960 ELECTRONIC GNG SWITCHES Wendell S. Miller, The Viking Bldg., Suite 201, 9606 v Santa MonicafBlvd., Beverly Hills, Calif.

Fixed Dec. zo, 1954, ser. No. 416,127

1 claim. (01340-166) l This invention relates to improved electronic gang switches, that is, devices adapted upon a single actuation to simultaneously switch a number of circuits between predetermined conditions. Certain features of the switches are claimed in my copending application Ser. No. 476,126, now U.S. Patent No. 2,856,596, entitled Magnetic Control Systems, and filed of even date herewith.

, In prior electronic gang switches, it has been necessary to employ numerous vacuum tubes to perform the switching function, and as a result such switches have been both expensive and unduly susceptible to failure in use. The general object of the present invention is to provide an improved type of gang switch in which no vacuum tubes are utilized in the switching operation, and which is therefore considerably simpler and more reliable than prior switches. At the same time, a switch constructed in accordance with the invention is extremely rapid in operation, and may therefore be employed in high speed computing devices requiring the performance of a large number of switching and control operations in extremely short intervals of time.

Structurally, a switch embodying the invention includes a matrix or pattern of magnetizable cores, preferably taking the form of a number of separate and spaced rings of magnetic material. These cores are desirably of a material having high loss essentially rectangular hysteresis loops, so that a magnetizing current or force has little. effect on the magnetic state of the core until the current reaches a predetermined critical value, at which time the current becomes effective to suddenly shift the magnetic state of the core, say from a negative driven state to a positive driven state. Such magnetization of the cores is effected by a plurality of sets of conductors, which may extend in two different directions (typically perpendicula-rly to form X and Y coordinates) past or through the different cores. The cores, and the energizing pulses impressed lon the conductors, may be so chosen that a single pulse on a single conductor adjacent one of the cores is not capable of shifting the magnetic state of the core, while a predetermined plurality of simultaneous pulses on more than one of the conductors will so shift the core.

A particular feature of novelty of the invention resides in the provision of a unique read-out or output arrangement, which allows signals to be read from the individual cores simultaneously, and which therefore allows for the functioning of the device as a multipole or gang switch. Specifically, I provide a plurality of separate read-out lines which are positioned to pick up signals from different ones of the cores, and are separately connectable to different read-out circuits to simultaneously control or switch those different circuits upon a single actuation of the device. The actual switching of the assembly from one condition to another may be effected by de-energizing one wire of one of the two sets, and energizing another of the wires. T he energization of the second set of A further feature of the invention resides in the provision of a novel type of switching pulse, -which serves to automatically lreturn the cores to a predetermined normal magnetic condition after each energization, and also serves to amplify or maximize the intensity of the output pulses. These results are attained by employing a switching pulse having both anVA.C. componentand a D.C, component, theformer preferably being greater than the latter. The total of these components may be such as to normally maintain each core in a predetermined state,- say a negatively driven state, and yet permit a relatively small auxiliary pulse on a second line to trip or actuate the core to the opposite or positive state. After the auxiliary pulse is terminated, the combined A.C. and D.C. pulse returns the core to its normal state.

The above and other features and objects of the present invention will be better understood from the following detailed description of the typical embodiments illustrated in the accompanying drawings in which:

Fig. l is a schematic representation of an electronic gang switch constructed in accordance with the invention; and

Fig. 2 is a representation of the hysteresis curve of the magnetic cores utilized in the Fig. l switch arrangement, together with representations of the pulses which are impressed on the signal input wires of Fig. l for causing a shift .in the magnetic state of the cores.

Referring first to Fig. 1, I have there shown an electronic gang switch embodying the invention, which switch may be a portion of a relatively complex electronic computer system. To simplify and clarify the present disclosure, I have included in Fig. 1 only as much of the overall computer mechanism as is necessary to an understanding of the functioning of the novel gang switch structure. The electronic gang switch of Fig. 1 comprises a number of individual cores 10 of magnetizable or other control circuit.

material, preferably taking the form of small rings, as shown. These cores may be arranged in both horizontal and vertical rows, there typically being shown three horizontal rows and four vertical rows. The cores are magnetically energizable by two sets of pulse input lines which extend through the cores, the first set of lines or conductors being designated X1, X2 and X3, and the second set of conductors being designated Y1, Y2, Y3 and Y4. The magnetic signals impressed on cores 10 are read from the cores by individual read-out lines 11, each extending through only one of the cores.

Each of the read-out lines 11 is connected to an individual read-out circuit 12, which is adapted to produce a desired indication, or to effect a desired control action, in response to energization of the associated core 10. For simplicity of illustration, only three of the read-out circuits 12 have |been represented in Fig. l, but it will of course be understood that other similar circuits are provided for the other read-out lines 11. The cores 10 and the various input and read-out conductors are of course mounted to a suitable `frame (not shown) adapted to hold these various elements in the illustrated relative positions. The cores and wires are spaced apart sufficiently far to assure that no core will be affected substantially by the magnetic field from any other core, or by the field of any wire other than the particular wires which pass through that core. Similarly, no read-out wire 1'1 is affected magnetically by any element other Vthan the particular core through which it passes, and the input wires which also pass through that same core.

Each of the vertical conductors Y1, Y2, Y3 and Y4 extends through the cores 10 of only one of the vertical rows of cores. Similarly, each of the conductors X11, X2 and X3 extends through only the cores in one of the horizontal rows. Consequently, there extendsthrough each of the cores 10 both an X wire and a Y wire, which wires are in iiux linkage relation with the core, and therefore control its magnetic state.

The various X and Y input lines are supplied with controlled electric pulses, which are utilized to shift the various cores from one magnetic state to the opposite magnetic state. The cores and pulse sources are so designed that no single pulse from any one of the X or Y wires is of suiiicient strength to shift the core magnetically to its actuated condition, and yet two simultaneous pulses, in both the X wire and the Y wire passing through a particular core are sufficient to actuate the core. In this connection, attention is called to Fig. 2, which represents at 13 the preferred hysteresis loop of the cores 10. As seen in that figure, the cores are preferably formed of an electrically magnetizable material whose hysteresis loop is of the high loss type, and has a rather sharp bend or knee at points 14 and 14a. The hysteresis loop 13 is desirably of the illustrated rectangular configuration, having substantially horizontal bottom and top sides 15 and 16, and having two substantially vertical sides 17 and 18. A suitable material 4for the purpose is the product sold by General Ceramics under the designation Ferrarnic Sl or S3.

Preferably, each of the cores 10 is normally maintained magnetized in one direction, typicallythe direction of the magnetic state represented by the lower line 15 in the Fig. 2 hysteresis loop. To simplify the discussion, this magnetic state represented at 15 may be termed a negative magnetic state, or negatively driven state, while the condition represented by upper line 16 may be called a positive state. As will be understood, in order to actuate any one of the cores from negative state 15 to positive state 16, the magnetizing force H or magnetizing current must be sufficiently great to drive the core past bend 14 of the hysteresis loop, following which the core abruptly changes to the positive state represented at 16. The pulses 4fed to lines X1, X2 and X3 are insucient by themselves to cause the core to pass bend 14, but will do so in combination with pulses from the Y lines. Unidirectional or direct current pulses are supplied separately to the various Y lines by individual pulse sources 19, which may typically be other portions of the same overall computng mechanism of which the illustrated cores and wires are a part. The positive side of each pulse source 19 may be connected to the upper end of the corresponding Y wire, while the negative side of the pulse source may be connected to the lower end ofthe same wire. To simplify the drawing, the negative line has only been shown in one of the Y circuits of Fig. l, but it will be understood that corresponding negative lines are actually provided in each of the other Y circuits.

The three X lines are adapted to serve as control or switch lines and may be selectively connectiible to a common pulse source. This pulse source preferably supplies to the X lines a composite electrical current or pulse, having an AC. component biased by a DC. component. The A.C. component may be supplied by an oscillator 20, which is connected in series with a battery 21 for supplying the DC. component. The three X lines are individually connectible to power sources and 21 by means of control switches 22, 23 and 24, which may be mechanical switches, electronic switches, or other switching means. A timer 25 may be provided for timing the pulses from sources 19, the timer desirably being operated by or in accordance with the current fluctuations of oscillator 26, to synchronize the X wire pulses with the Y wire pulses in a manner to be discussed more specifically at alater point.

ln Fig. 2, I have represented at 26 the pulses which are fed to the X lines by oscillator 20 and battery 21, and i have represented at 27 the D.C. pulses which are fed to the Y lines by sources 19. The pulses 26 have a D.C component 28 supplied by battery 21, which cornponentshiftstlie center of the A.C. cycle leftward to the point 29 in Fig. 2. The A.C. component is preferably greater than the D.C. component. The combined and D C. components of pulses 26 provide a magnetlzing force H which tluctuates lbetween a left limit 30 and a right limit 31. When pulse 26 reaches its left limit 30, the magnetizing force H is sufficiently `great in a negative direction to cause the associated core 10 to magnetically pass upper ibend 14a of the hysteresis loop, and thus actuate the core to its negative driven state represented at 15. This is true even though there may be no pulse 27 supplied by the associated Y wire. However, the other extremity of pulse 26 in Fig. 2, that is, the right extremity represented at 31, does not xprovide a suflicientmagnetizing force H to pass bend 14 of the hysteresis loop and thus actuate the core to its positive state 16. Consequently, unless a pulse 27 is supplied by the corresponding Y wire, the core is not actuated to its positive state, even though an X pulse 26 is present. The X pulses thus serve to maintain each core 10 in its negative state 15, until a pulse 27 is supplied on a Y wire, at which time the combination of pulses actuates the core to its positive state 16, following which the pulse acts to return the core to its negative driven state 15 (assuming that the Y pulse 27 has been terminated).

In actual operation, the X and Y pulses are of course fed to the various input lines in very rapid succession, and in accurately timed relation, and act to control the energization of read-out circuits 12 just as rapidly. The method of producing these pulses may vary in different types of computers or other devices, and any of the various known types `of pulse control systems may serve the function of the elements generally representsd as pulse sources v19 in Fig. 1. The timer 25 is provided for controlling the exact timing of the D.C. pulses supplied by sources 19, with respect to the A.C. pulses supplied by oscillator 20. When one of the switches 22, 23 or 24 is closed, the D.C. biased A C. pulses 26 are supplied repetitively to the corresponding X line. If any of the sources 19 are then in a state for supplying D.C. pulses 27 to the corresponding Y lines, timer 25 so synchronizes these sources 19 with oscillator 29 las to assure that each pulse 2.7 adds to or supplements the correspondingly directed portion of the oscillating pulses 26. That is, pulse 27 occurs while the pulse 26 is in a direction such that the magnetizing effect of pulse 26 is in the same direction as pulse 27. Thus, these two combined pulses cause.- the core to shift to its positive state 16, following which the pulse 27 terminates as pulse 26 reverses to a negative state, so that pulse 26 then returns the core to its negative state 15 until the next pulse 27 occurs. Such actuation of a core 1t) to its positive state 16 creates a magnetic field in the vicinity of that core which induces an electrical current in the corresponding read-out line 11, to thus energize read-out circuit 12 which acts to thus indicate or respond to the core actuation.

It will be understood from the above that the Fig. l arrangement will serve very effectively as a gang switch, in which actuation of any one of the three switches 22, 23 or 24 will act to simultaneously contro-l or switcha number of different electrical circuits. The number of circuits associated with each switch 22, 23 or 24 and the corresponding X line may of course be far more than the typically illustrated four circuits. If, for example, switch 22 is closed, the resultant energization of line X1 with repeating D C. biased A.C. pulses of the type represented at 26 in Fig. 2, causes the cores `10, through which line X1 passes, to become responsive to pulses 27 supplied by sources 19. lf any one of these sources then supplies a D.C. pulse 27, the combination of that pulse with the pulse 26 in line X1 will cause a Vcorresponding one ofthe upper four cores 10 to shift momentarily to its positive state 16, and then return to its negative state. This shift is indicated lby a change in condition of the corresponding read-out circuit 12, so that four or more read-out Giruts 12.(asso'ciated with the upper four cores respectively) respond to the control switch 22. Any of the other circuits may be correspondingly closed, to correspondingly actuate the various read-out circuits associated therewith.

I claim:

An electronic gang switch comprising a series of separately energizable input conductors, a series of separately energizable control conductors, a matrix of individual cores of magnetizable material each positioned in liux linkage relation to one of said input conductors and one of said control conductors, said cores having substantially rectangular hysteresis loops with two stable states of magnetization, a control signal source supplying electrical signals to said control conductors selectively and including a first source supplying an A.C. wave and a second source super-imposing a D.C. component on said A.C. wave to form a composite control signal having both A C. and D.C. components, said A.C. wave having `a peak current of greater amplitude than the current amplitude of said D.C. component, said composite control signal producing a fluctuating magnetizing force which by itself is strong enough in a rst direction to actuate said cores to a predetermined normal magnetic state but is not strong enough in the opposite direction to actuate said cores to the opposite magnetic state, a plurality of electrical input signal sources supplying D.C. input pulses to a plurality of said input conductors simultaneously and actuable to do so in any of several different possible combinations of the diierent input conductors, said D.C. input pulses being in a direction tending to magnetize the cores oppositely from said D.C. components of the control signal and being of an intensity serving with said composite control signal to energize the cores to said opposite magnetic state, synchronizing means for timing said D.C. input pulses to occur simultaneously with the portion of the A.C. cycle of said control signal which is in the sarne direction as said input pulses, said cores and conductors being arranged in a pattern such that if a first control conductor is energized in combination with a selected group of the input conductors a rst group of cores will be simultaneously actuated to said opposite magnetic state, and if a dilferent control conductor is energized in combination with the same group of input conductors a different group of cores will be simultaneously actuated to said opposite state, a plurality of individual read-out conductors passing in iiuX-linkage relation with diierent cores of said matrix, and a plurality of different read-out circuits actuable separately by said different read-out conductors respectively, said readout conductors being arranged in a pattern such that a predetermined plurality of the diiierent read-out circuits will be simultaneously actuated when a particular control conductor yand a particular group of input conductors are energized.

References Cited in the le of this patent UNITED STATES PATENTS Rosenberg Oct. 5, 1954 Saltz Oct. 5, 1954 Electronics, April 1953, pp. 146-149. Electronic Engineering, May 1954, pp. 192-199.

Images