* GB785894 (A)

Description: GB785894 (A) ? 1957-11-06

Improvements in or relating to electrical circuits of the kind including anelectro-magnetic device

Description of GB785894 (A)

PATENT SPECIFICATION 785,894 Date of filing Complete Specification: July 3, 1956. Application l ate: June 13, 1955 No 16999 /55. Complete Specification Published Nov 6, 1957. Index at Acceptance:-Classes 40 ( 1), RUB 17 (D: E); and 40 ( 4), XBP 6. International Classification:-008 c H 04 m. COMPLETE SPECIFICATION. Improvements in or relating to Electrical Circuits of the Kind including an Electro-Magnetic Device. We, THIE GENERAL ELECTRIC COMPANY LIMITED, of Magnet House, Kingsway, London, W C 2, a British Company, and NORMAN CAROL SMART, of The General Electric Company Limited, Telephone Works, Coventry, a British Subject, do hereby declare the invention, for which we pray that a patent may be granted to us, and the method by which it is to be performed, to be particularly described in and by the following statement:- This invention relates to electrical circuits of the kind including an electromagnetic device, for example a step-by-step switch that is adapted to be operated by electric impulses. If an electromagnetic step-by-step switch is operated over line wires from a remote source of electric impulses, the ohmic resistance and hence the length of the line wires must usually be restricted to such a value that the current flow thereover during an impulse is sufficient to operate the switch. In many applications, such as systems of remote control and supervision and telephone systems, where the line wires may be of considerable length this restriction is sometimes found to be too severe, particularly when the switch is a uniselector or a minor

switch One method of alleviating this restriction is to employ a switch requiring less current to effect operation and to this end it is known to employ special switches, so-called relay switches An alternate method is to make more effective use of the energy available from the impulse source. It is one object of the present invention to provide an electrical circuit of the kind specified that does make more effective use of the available energy. lPrice 3 s 6 d l It is another object of the present invention to provide an improved electrical circuit which includes an electromagnetic stepby-step switch that is adapted to be operated 45 by electric impulses supplied over line wires from a remote source, the circuit being such that a larger value of ohmic resistance is permissible for the line wires than heretofore without employing a switch having 50 special operating characteristics. According to the present invention, an electrical circuit of the kind specified includes capacitance and means responsive to the supply of electrical impulses to the cir 55 cuit to apply each impulse to the capacitance so that at least part of the electrical energy of the impulse is stored by the capacitance, the said means being adapted to act upon the termination of each impulse 60 so as to connect the capacitance in a closed circuit with the operating winding of the electromagnetic device whereby the capacitance discharges through the operating winding to operate the said device and the 65 arrangement being such that the electrical power supplied to the operating winding upon the termination of any one of the impulses is greater than would have been the case if the impulse had been supplied 70 directly to the operating winding. Two embodiments of electrical circuits in accordance with the present invention will now be described, by way of example, with reference to the two Figures of the accom 75 panying diagrammatic drawing, in which: Figure 1 shows an electrical circuit in which the electromagnetic device is a uniselector switch; and Figure 2 shows an electrical circuit in 80 which the electromagnetic device is a minor switch. 785,894 Each electrical circuit to be hereinafter described is particularly suited to electrical systems of remote control and supervision where it may be employed at a sub-station to receive control impulses from the main station and to set the electromagnetic stepby-step switch accordingly Each circuit may also be employed in automatic telephone systems, as, for example, in connection with a multi-party telephone line to afford discrimination between the parties connected to that line Thus each party may be particularised by a different position of the wipers of the electromagnetic

step-bystep switch, a party being selected by the application of the appropriate number of impulses to the line wires to operate that switch accordingly. The electrical circuit shown in Figure 1 is provided with two input terminals 1 and 2 whereby it is arranged to be connected to line wires (not shown) and thence to a remote source of electrical impulses (not shown) An electromagnetic relay A is adapted to respond quickly to electric impulses supplied to the terminals 1 and 2 and has a high resistance winding 3 which is permanently connected across the input terminals The relay A has a set Al of changeover contacts, the operating contact 4 of the set being connected to the input terminal 1 by way of a capacitor 5 which is of high value The fixed contact 6, with which the operating contact 4 is in engagement when the relay A is non-operated, is also connected to the terminal 1, this connection being by way of the operating winding 7 of the uniselector switch 8 The other fixed contact 9 is connected to the input terminal 2 by way of a rectifier element 10 the function of which will be described hereinafter. During operation the input terminals 1 and 2 are connected to the line wires (not shown) Upon the supply of an electric impulse over the line wires the relay A operates For the circuit shown, an impulse consists of negative voltage with respect to earth applied to the line wire connected to the terminal 2 and earth potential applied to the line wire connected to the terminal 1. Upon the relay A operating its contacts 4 and 9 connect the capacitor 5 in series with the rectifier element 10 between the terminals 1 and 2 The rectifier element 10 is -55 so connected as to be conductive to the current flow which results from the potential difference between the line wires during an impulse The capacitor 5 therefore charges for the remainder of the impulse, the relay A remaining energised By having a high resistance for the winding 3 the current required to keep the relay A energised during an impulse is kept to a low value so that the majority of the electrical energy of an impulse is stored by the capacitor 5. Upon the termination of the impulse, the capacitor 5 is prevented from discharging through the winding 3 or to the line by the rectifier element 10 The relay A therefore releases so that the operating contact 4 70 restores to the contact 6 and connects the capacitor 5 in a closed circuit with the switch operating winding 7 The capacitor 5 discharges through the winding 7 so that the switch 8 is operated, its one or more wipers 75 such as 11 being each stepped one contact position around the associated contact bank such as 12 Each subsequent impulse of a series of impulses supplied to the input terminals 1 and 2 results in a repetition of So the cycle described above, the switch

wipers 11 and the like being stepped accordingly. It will be appreciated that of the current flow resulting from an impulse the component which charges the capacitor 5 de 85 creases exponentially Consequently the voltage drop across the line wires resulting from that component of the current also decreases in a similar manner so that the voltage appearing between the terminals 1 90 and 2 rises In this manner it is arranged that the power supplied to the switch winding 7 by the capacitor 5 when the relay A releases is appreciably greater than would have been the case if the relay A had been 95 omitted and the switch winding 7 connected directly between the input terminals 1 and 2 In other words the present invention makes more effective use of the available impulse energy supplied to the input 100 terminals. In a typical example, the line wires (not shown) may have a loop resistance of 1,000 ohms and the capacitor 5 may have a value of 80 micro-farads while the impulses sup 105 plied to the line wires by the source (not shown) may have an amplitude of 100 volts and a duration of 75 milli-seconds, there being an interval of 25 milli-seconds between succeeding impulses of the series 110 It will be necessary to home the switch wipers 11 and the like after use Since the switch 8 is a uniselector this can be effected by supplying further impulses to the input terminals 1 and 2 115 In the circuit shown in Figure 2 the electro-magnetic device is a minor switch 13 having a release magnet 14 which, when energised, permits the switch wipers such as 15 to be restored mechanically to their 120 home position That part of the circuit which is concerned with the stepping of the switch wipers such as 15 is substantially as previously described and corresponding circuit elements carry the same references as 125 in Figure 1 An additional relay B is provided to facilitate the operation of the release magnet 14 in response to a further impulse supplied to the input terminals 1 and 2 after a series of impulses supplied to 130 adapted to act upon the termination of each impulse so as to connect the capacitance in a closed circuit with the operating winding 65 of the electromagnetic device whereby the capacitance discharges through the operating winding to operate the said device and the arrangement being such that the electrical power supplied to the operating wind 70 ing upon the termination of any one of the impulses is greater than would have been the case if the impulse had been supplied directly to the operating winding. 2 An electrical circuit according to 75

* Sitemap * Accessibility * Legal notice

* Terms of use * Last updated: 08.04.2015 * Worldwide Database * 5.8.23.4; 93p

* GB785895 (A)

Description: GB785895 (A) ? 1957-11-06

Improvements in and relating to squirrel cage motors

Description of GB785895 (A)

PATENT SPECFICATION Inventor: -WALTER HILL. i Date of filing Complete Specification: June 22, 1956. Application Date: June 21, 1955 No 17954 /55. Complete Specification Published: Nov 6, 1957. Index at Acceptance-Class 35, A( 6 B:1682). International Classificatioun:-HO 2 k. COMPLETE SPECIFICATION. Improvements in and relating to Squirrel Cage Motors. We, METROPOLITAN-VICKERS ELECTRICAL COMPANY LIMITED, a British Company, having its registered office at St Paul's Corner, 1-3 St Paul's Churchyard, London, E C 4, do hereby declare the invention, for which we pray that a patent may be granted to us, and the method by which it is to be performed, to be particularly described in and by the following statement: - This invention relates to squirrel cage motors particularly to their rotors, and aims at providing an improved construction which is relatively simple in manufacture and ensures a desirable operating characteristic. In connection with polyphase squirrel cage induction motors it has been known to make use of a current displacement phenomenon in order to obtain an operating characteristic affording high starting torque, low starting current, good full load efficiency, satisfactory power factor and reasonable overload capacity. The known designs are based on the fact that the frequency in the rotor circuit is equal to the supply frequency when the rotor is at

standstill, and falls as the motor runs up to full load speed, at which speed the frequency in the rotor circuit is a small fraction only of the supply frequency. Owing to this frequency change the rotor current is displaced from an outer rotor winding to an inner rotor winding as the motor speed increases and the desirable operating features mentioned above are at least partly obtainable by a double-cage or Boucherot type rotor which has a high resistance and low reactance in the outer cage forming the outer winding, and a low resistance but high reactance in the inner cage forming the inner winding, the current lPrice 3 s 6 d l being displaced from the outer to the inner cage as the rotor runs up from standstill to full speed. In order to ensure for a given overload capacity a high starting torque at low starting current, it is desirable to reduce the outer cage reactance to a minimum. It will be also understood that the inner cage resistance should be as low as possible in order to increase the full load efficiency. Both the outer cage reactance and the inner cage resistance can be reduced by increasing the number of rotor bars. It is also known that equality of the numbers of slots in the inner and outer cages affords optimal flux distribution and load losses, which are more favourable for instance than those obtaining with half the number of slots in the inner cage. The present invention relates to an improved constructon which makes it possible to provide a high number of inner and outer slots, while at the same time ensuring an inner cage resistance which is substantially lower than that of conventional double cage rotors With a modified type of conventional rotors, the inner winding comprises bars which are radially offset However, in this winding bars spaced approximately by a pole-pitch are connected in series employing suitably shaped end constructions which add appreciably to the total resistance of the inner cage circuit and this disadvantage is avoided by the construction of the present invention. The characteristic features and limitations of the various known forms of double cage rotors, as compared wth the invention will be best understood by reference to the drawings accompanying the Provisional Specification in which: 7855895 785,895 Figs 1 and 2 show double cage rotor cross-sections of two ordinary designs; Fig 3 shows the modified arrangement mentioned above; and a Fig 4 illustrates a preferred embodiment of the present invention. If the construction of Fig 1 is used the number of slots is limited as the rotor body becomes mechanically weak at the dimension A often before the flux is limited by the dimension B. The arrangement of Fig 2 permits a larger number of slots than Fig 1, the maximum flux being

here limited by the dimension D. Obviously the arrangement of Fig 3 affords a further increase in the number of rotor bars. However, equal currents must pass through the two groups of the radially offset bars defined by the slots, E, G and F, H respectively of the inner winding to ensure an operating characteristic which combines lowest starting currents with highest starting and over-load running torques In a conventional construction this equality of currents is obtained by the series connection of bars which are radially offset and approximately a pole-pitch apart, such as o U bars E and F and bars G and H. For such series connection suitably shaped end connectors must be nrovided and according to the present invention these shaped members can be dispensed with, 3 < whereby a lower inner cage resistance and therefore a higher full load efficiency are ensured Fig 4 shows a preferred form to achieve this effect. In this arrangement the number of rotor bars may be the same as that of a rotor according to Fig 3 However, the flow of equal currents through bars positioned at different radial distances, such as bars K and M is now ensured by equal impedances of the bars. This is achieved by using bars of equal resistance and proportioning the slot opening width wk and length 1 k of slot K and the slot opening width tam and length I of slot M so in relation to each other that 1 k lm Wk Wm Thus both sets of inner winding bars have the same impedance, equal currents will flow, and both sets of bars of the inner winding can be connected, by brazing for instance, to short-circuiting rings at both ends of the core. It will be understood that the outer cage slot configuration, which forms no part of this invention, can be chosen as is known in the art For instance all these slots can be uniform with identical widths and depths of slot openings.

* Sitemap * Accessibility * Legal notice * Terms of use * Last updated: 08.04.2015 * Worldwide Database * 5.8.23.4; 93p

* GB785896 (A)

Description: GB785896 (A) ? 1957-11-06

Corner fittings for use in the construction of sheet metal boxes

Description of GB785896 (A)

PATENT SPECIFICATION Inventor: -SIDNEY ROLAND LOWE. Date of filing Complete Specification: June 26, 1956. Application Date: July 4, 1955 L No 19231/55. Complete Specification Published: Nov 6, 1957. Index at Acceptance:-Classes 18, A( 4 f 1: 5); 38 i 1), E 28; and 38 ( 5), B 2 B 7. International Classification:-B 65 d H On H 02 f. COMPLETE SPECIFICATION. Corner Fittings for Use in the Construction of Sheet Metal Boxes. We, MIDLAND ELECTRIC MANUFACTURING COMPANY LIMITED, of M E M Works, Reddings Lane, Tyseley, in the City of Birmingham 11, a British Company, do hereby declare the invention, for which we pray that a patent may be granted to us, and the method by which it is to be performed, to be particularly described in and by the following statement: This invention relates to metal boxes constructed from metal sheets which are secured together by internal corner gussets, and particularly wall boxes for containing electric switches, fuses or other electric connecting means, the boxes being provided with external lugs through which screws can be inserted for securing the boxes in position. The object of the present invention is to provide the gussets and lugs in a more convenient form than heretofore, which enables the lugs to be conveniently disposed in any desired positions relatively to the box. A corner fitting in accordance with the invention comprises a gusset adapted to occupy a corner of the box and having an end flange which extends between the side members of the gusset, a lug adapted for attachment to the flange by a screw, and means for interengaging the lug and flange consisting of a projection on one of the parts and a series of notches on the other part. In the accompanying drawings. Figure 1 is a diagrammatic view of the base of a box Drovided with corner fittings in accordance with the invention.

Figures 2 and 3 are respectively end and side elevations illustrating one of the corner fittings. Referring to the drawings, the rectangular sheet metal box there shown consists of at least four side pieces a made from sheet metal and (optionally) a sheet metal base b For such a box four corner fittings 45 are required, each comprising a gusset c, and a lug d Each gusset is made of rightangle section and has formed on at least one of its ends a flange e which extends between the side members of the gusset, and in the 50 flange is formed a screw threaded hole for reception of a lug attachment screw Also on this end of the gusset is formed a short projection f which extends beyond the side of the box to which it is required to attach 55 the lug. Each lug has formed in it two holes g, h, the hole g serving to receive a screw i for securing the lug to the flange e on the gusset, and the hole h serving to receive a screw or 60 nail for securing the box to a wall or other support Preferably the lug is provided with a recess j to accommodate the head of the screw i. It is required that the lug shall be secur 65 able to the box in any of several positions, and to meet this requirement one face of the lug has formed in it a series of suitably disposed notches k any of which can be engaged with the projection f on the gusset 70 flange, the lug being held in the selected position by the screw i which attaches it to the flange. Figure 1 shows the lugs extending parallel with the side edge of the box, as do also 75 Figures 2 and 3 Alternatively the lugs may extend outwardly at right angles to these sides When it is desired that the securing nail or screw inserted through the hole h shall lie within the box, the base of the box 80 is provided with holes m In this case each lug can be moved to a position in which its hole h coincides with a hole m, the 785,896 nail or screw being then inserted through both of the associated holes ra h. In an alternative construction the notches k are formed in the flange of the gusset, and the projection f is formed on the lug. By the use in a sheet metal box construction, of corner fittings as above described, the provision of lugs which are required to occupy selectable positions on the box is effected in a simple and convenient manner.

* Sitemap * Accessibility * Legal notice * Terms of use * Last updated: 08.04.2015 * Worldwide Database

* 5.8.23.4; 93p

* GB785897 (A)

Description: GB785897 (A) ? 1957-11-06

Improvements in or relating to electrical information storage and read-outdevices

Description of GB785897 (A)

A high quality text as facsimile in your desired language may be available amongst the following family members:

BE541151 (A) DE1041535 (B) FR1132925 (A) US2732542 (A) BE541151 (A) DE1041535 (B) FR1132925 (A) US2732542 (A) less Translate this text into Tooltip

[85][(1)__Select language] Translate this text into

The EPO does not accept any responsibility for the accuracy of data and information originating from other authorities than the EPO; in particular, the EPO does not guarantee that they are complete, up-to-date or fit for specific purposes.

PATENT SPECIFICATION 785 897 0 Date of Application and filing Complete Specification: Sept 2,1955. No 25277 /55. Application made in United States of America on Sept 13, 1954. Complete Specification Published: Nov 6, 1957. Index at acceptance:-Classes 40 ( 4), B; and 106 ( 1), C( 1 D; 2 H; 2 K: 6). International Classification:-G 06 f. COMPLETE SPECIFICATION Improvements in or relating to Electrical Information 7 Storage and Read-Out Devices We, WESTERN ELECTRIC COMPANY, INCOPPORATED, of 195, Broadway, New York City, New York State, United States of America, a Corporation of the State of New York, United States of America, do

hereby declare the invention for which we pray that a patent may be granted to us, and the method by which it is to be performed, to be particularly described in and by the following statement: - This invention relates to electrical information storage and read-out devices of the type described in British Patent No 760,307, that is to say comprising an array of magnetic cores formed by areas defined by holes in a single sheet of magnetic material and through which holes wires are threaded to intersect the cores. In magnetic core matrices, information is stored by switching or setting the magnetic state of a single core in the array during a particular time interval; this is done by pulsing selection wires which all thread or intersect only that one core Subsequently the information is read out by sensing that core; this is done by applying sensing pulses to the particular selection wires to switch the state of the core again, if information has been stored therein, and cause an output pulse to appear on a read-out or output wire that also threads the core As only one core is sensed at a time and thus any information appearing on the read-out or output wire must be applicable to the core that is being sensed, it has become usual to have but a single readout wire thread or intersect all the cores in the array. It is possible, however, to have the information appearing on the read-out wire -40 erroneous Let us consider first that the pulses applied to the selection wires threading others than the selected core are sufficient to affect the state of those cores; as pointed out in British Patent No 760,307, these pulses, particularly if applied repetitively, may lPrice 3 s6 d l change the state of the core, even though the magnetic state is not completely switched, so that the information stored therein cannot be read out properly As described in the abovementioned patent, the effects of these pulses on the individual cores may rendered negligible, so that masking or destruction of the information stored in a core does not occur, by threading each core with a number of wires to each of which is applied an equal part of the setting current desired Further, the sets of selection wires may be so threaded through the cores that no two selection wires jointly thread more than one core in the array In this manner, the disturbing current applied to any core is limited to just 1/p the setting, current, considering there to be p se:s of selection wires As more fully described in the above-mentioned patent, this can be attained in an N by N matrix if a first group of wires thread each core in a row in the a coordinate direction, a second group of wires thread each core in a column in the b coordinate direction, and a plurality of other groups of wires each thread the cores The wires of the other groups have slopes such that for any two wires whose slopes

are defined by the pairs (ab) and (a 1,b 1,), the difference (ab'-ba') has no common factor with N and is not zero If the array of magnetic cores is formed by areas defined by holes in a sheet of magnetic material, as described in the above-mentioned patent, there is another criterion for acceptable wires in the storage matrix, namely, that the interval in the a direction plus the interval in the b direction between successive cores threaded by a selection wire is constant and odd, except where an edge of the matrix is traversed between successive cores. The information stored in the individual cores in the matrix will not be destroyed on successive storing and sensing of other cores in the array However, even though the information stored in any one core will not be destroyed by these disturbing pulses applied to the other cores in the array, the error pulses induced in the single read-out wire threading all the cores due to these disturbing pulses themselves may be sufficient to mask the output pulse read from the sensed core The first source of erroneous output pulses, discussed above, is due to change in the magnetic state of the core; this source of erroneous pulses is due to the coupling between the selection and read-out wires. As the number of groups of selection wires is increased to prevent destruction of the information stored in any one core, the total number of cores to which disturbing pulses are applied when a single core is sensed is increased, and while the effect on the single read-out wire of the disturbing pulse at one of these other cores is very small, the summation of all these small effects may be of such a value as to mask the proper signal on the read-out wire from the sensed core. The present invention provides 'an electrical information storage and read-out device of the type set forth above having an N by n array of magnetic cores, a plurality of selection wires intersecting each of the cores for storing information therein and sensing the stored information and of which the wires intersecting any one of the cores all intersect that core in the same direction, and a single read-out wire intersecting all of the cores and in which an output pulse is produced when the information stored in any one of the cores is sensed, wherein, for the purpose of reducing the masking effect on the output pulse of the error pulses which are induced in the read-out wire due to the passage of the selection wires through other cores of the array, n+ 1 the read-out wire is caused to intersect 2 n+l of the cores, or the integral part of of 2 the cores, in the first row of the array of magnetic cores, in the same direction as the selection wires passing through such cores, and to intersect the remaining cores in the first row in a direction opposite to the direction of the

selection wires therethrough, the directional threading pattern for the read-out wire which is thus established for the first row of the array being repeated for each succeeding row of the array, but successively shifted by S cores, where S is a prime number relative to n. The manner in which the above directional pattern for the read-out wire is arrived at will be apparent hereinafter, it being advantageous to define first certain terms to enable us to specify the criteria governing the method of threading of the read-out wire through the array of magnetic cores which are required in order to attain substantial cancellation of these error pulses in the read-out wire This can best be done with reference to the accompanying drawings, in which: Fig 1 is a partial schematic representation 65 of a 7 x 7 magnetic core matrix employing four sets of selection wires, the selection wires for only one core in the matrix being depicted and being threaded through the matrix in the manner described in the above-mentioned 70 British patent; Fig 2 is a schematic representation of the threading of a single read-out wire through each of the cores of the matrix of Fig 1, likewise in the manner described in the 75 above-mentioned patent; and Fig 3 is a schematic representation of the threading of a single read-out wire through each of the cores of the matrix of Fig 1 in accordance with one illustrative embodiment 80 of this invention. In small magnetic core arrays one can advantageously employ what we shall refer to as a standard output arrangement in which the read-out wire threads the cores in a 85 serpentine fashion, alternately of the same and opposite directions as the selection wires. With but two coordinate selection wires one can easily see which are the non-selected cores to which disturbing pulses are applied and in 90 what relative directions the selection and read-out wires intersect these cores When the number of cores in the matrix and the number of selection wires increase this cannot easily be seen In the above-mentioned 95 British patent a 5 x 5 matrix with four sets of selection wires is disclosed, employing the standard serpentine output arrangement. This complexity can best be seen if we consider four selection wires threading but 100 one core in a 7 x 7 matrix of the type disclosed in the above-mentioned patent This is depicted in Fig 1 wherein the matrix is assumed to be of the type disclosed in the above-mentioned patent with the cores pro 105 vided by a sheet 10 of magnetic material. The magnetic cores 12 through which only one of the selection wires 15 and 16 pass are in darker outline than those not involved in the

selection or sensing of the single core 13 110 As can be seen of the forty-nine cores in the array, twenty-four cores are intersected by a single selection wire 15 or 16 having a disturbing pulse thereon in addition, of course, to the selected core; further there 115 would be more non-selected but disturbed cores if additional selection wires were employed. To understand how, in this embodiment, the summation of the error pulses on the 120 single read-out wire is caused to be minimal by means of the invention, regardless of the selected core and therefore of the particular disturbed cores, it will be advisable to introduce several terms and notations In this way 125 the applicability of the invention to the 785,897 utilized The output wire 18 is connected to a detector circuit 19 and we will consider that it enters the matrix at the first core on the left of the first row, as indicated by the arrow. The output array Q for this arrangement of the output wire 18 can therefore be defined as follows: general case of an N by N matrix having p selection wires can more readily be seen The three terms to be defined first are the selection array S, the output array Q, and the product array P Let us define the binary quantities c" 1 " and " O " so that a " 1 " means that the wire is being threaded through the sheet from on top, considering the cores to be threaded in succession from the selection pulse sources to ground; thus a core intersected by a wire into the paper in Fig 1 will have a value " 1 " Similarly a core threaded by a wire coming up through the sheet 10 or out of the paper will have a value " O " The selection array S defines in which direction the selection wires 15 and 16 are threaded in each core in the array In the arrangement of selection wires disclosed in the above-mentioned patent and repeated, for one core, in Fig 1 of this drawing, the selection wires all thread any one core in the same direction and these selection wires include coordinate wires 15 which thread successive cores in the rows and columns of the array. As these coordinate wires 15 must thread the cores alternately, i e, entering the sheet at one core and emerging at the next, etc, the selection array itself will comprise alternate " l's " and " O 's " Accordingly, we can define the selection array as follows, assuming that the selection wires are threaded into the first core on the left of the first row of the matrix: 101 101 8 = O 1 O 101 101 0101 1010 0101 1010 0101 1010 0101 ( 1) The position of the binary " 1 " or " O " in the array is the same as the position of the core in the matrix Any pierced matrix of magnetic cores, of the type described in the above-mentioned patent, will have a selection array of this type.

Just as an array is associated with the direction of the intersection of the selection wires with the cores of the matrix, so another array, termed the output array Q, is associated with the direction of the output wire through each core Again a " 1 " in the array means the output or read-out wire is threaded into the core and a " O " that it is threaded out of the core As an example let us consider the standard output wire arrangement depicted in Fig 2 In this serpentine wiring arrangement the output wire 18 is arranged alternately to thread the cores in the same and opposite directions as the selection wires; as pointed out above this would effectively cancel the induced error pulses on the read-out wire due to the disturbed pulses at the non-set cores if only the coordinate selection wires 15 were 1100 0011 Q= 1 0 0 1 1101 011 001 011 ( 2) There remains to be defined now the product array P This can be defined mathematically by considering that the elements of the S and Q arrays are respectively sij and qis. Then the product array P comprises the elements pij defined by the expression pij-= siq 1 j + S 1 jq'jj ( 3) where the prime denotes the binary inverse. Or we can define the elements of the product array in these terms: if the selection wires and the read-out wires thread a core in the same direction, the element of the product array corresponding to that core in the matrix is " 1 " and if in opposite directions the element of the array corresponding to that core in the matrix is a zero. Accordingly, if the selection array for Fig. 1 and the output array for Fig 2 are utilized in a core matrix the product array would be 1001100 1100111 0110001 P= O O 111 O O 1000110 1110011 0011001 ( 4) 100 As a " 1 " in the product array indicates that the selection and output wires intersect that 105 core in the same direction and an " O " in the opposite direction, the " l's " and " O 's " in the product array P denote the relative polarities of the masking or error pulses in the single output wire due to the various cores in the 110 matrix If a given storage location is chosen, as in Fig 1, then the number of " I's " in P on the various selection wires minus the number of " O 's," exclusive of the selected core, indicates the number of unbalanced 115 error pulses due to the non-selected cores. It is not a simple matter, particularly as the number p of sets of selection wires and the size N of the matrix become large, to see which are the disturbed cores when a par 120 ticular core is selected To indicate the scope of this particular problem let us consider the 785,897 product array ( 4) above and the summation of the disturbing pulses when the core 13 is selected in the embodiment of Fig 1 To do this we shall rewrite the product array underlining those elements of the array corresponding to the heavily outlined cores of Fig 1,

namely, those cores 12 having applied thereto one disturbing pulse by a selection wire on sensing of the core 13 The product array is now: 1 O O 1 1 O O 1 1 O O 1 1 1 0 1 1 O 1 P=O O 1 1 100 ( 4 a) 151 O O l 1 1 O O i 6 o 1 00 o 1 The element in the product array of expression ( 4 a) corresponding to the selected core is marked by an X instead of by a " O " or " " As can be seen by a summation of the " O's " and " l's " there is an unbalance of six negative error pulses, i e, of six pulses produced at cores designated " O " in this for n= 5, the error signal per n= 7 n= 9 n= 11 n= 13 n= 15 n= 17 As can be seen the maximum unbalanced signal per set of selection wires approaches but does not exceed the possible maximum values of 3 5 and + 3 5 for larger matrices. For other arrays than the standard output array and what we have referred to as the standard output winding arrangement, the maximum unbalanced signal per set of selection wires may be considerably worse Further for other groups of selection wires it can be shown that the possible maximum unbalanced signal may be larger than 3 5 per set of selection wires. In practising the present invention the output wiring arrangement and the output array are chosen so that when the error pulses induced in the read-out wire 18 are summed together, for any group of pulsed selection wires 15 and 16, there is a maximum of cancellation Specifically, this is attained by threading the output wire through the cores of the matrix to intersect the cores in such a pattern that a particular product array is produced, which product array we shall refer to as a modular product array. The first row of a modular product array n+l n may consist of l l l's " and l-l " O's," 2 2 array that are not cancelled by positive pulses produced at cores designated " 1 " in this array This gives an unbalanced error signal of 6/4 or 1 5 per set of selection wires This is actually not too bad and can be utilized with cores defined by materials having good magnetic properties However, it can be shown that for the standard product array and thus for the standard output wiring arrangement as shown in Fig 2 the unbalanced error signal on the signal output wire per set of selection wires is between -3.5 and + 3 5, where the minus and plus indicate excess of "O's" and " l's" in the product array, respectively This relation is independent of the size N of the matrix for these particular selection wires; however, actually for smaller matrices the unbalanced signal per set of selection wires is smaller than the maximum predicted by the general relationship stated above Specifically, it can be shown that the maximum unbalanced error signal per set of selection wires for various size matrices having the four sets of selection wires shown in Fig 1 is given by the following table, where N of course indicates the size

of the matrix: set is between 1 and -2 -2 -2 -3 -2 2 5 + 1; + 2 + 2 + 2.5 + 3 + 3 + 2 arranged in that order, where the symbol lAl is to be read " the integral part of " A " The second row of the modular product array is generated by the first row, cyclically shifted by a number of elements S and specifically in one embodiment shifted to the left by one element of the array; the remaining rows are similarly formed Thus for the case of n= 7 the modular product array is, in this one embodiment, 111 1000 1110001 1100011 P= 1 O O O 1 1 1 0001111 0011110 0111100 ( 5) It can be shown that the range of the unbalanced signal per set of selection wires on the output wire due to disturbing pulses at non-selected cores for an output wire threaded through the matrix to give a modular product array of the above nature is between 1 and + 1, if N is even, and between 0 and + 2, if n is odd, regardless of the size of N and regardless of the number and particular sets of selection wires employed Comparing this 785,897 n+l pattern of l-l " l's " in the first row, which 2 pattern is shifted by one for each succeeding row In this embodiment we shall take the product array whose first row is given by the pattern range with that of the standard product array referred to above, it is readily apparent that its maximum limits are considerably smaller, which is particularly important and may in fact be crucial for very large matrices utilizing several hundred or thousands of cores in large scale memory devices. The modular product array need not start n+l with the l-l " l's " What is important to 2 minimize the possible unbalanced signal is that there be that number of "l's " in the first row of the array and that the pattern thus established in the first row be repeated in each succeeding row, but shifted by a specified number of elements, as discussed further below For example, other product arrays could have the pattern 1 1 O O 1 1 O O ( 6) in the first row, which pattern is shifted for each succeeding row For the cases where n is equal to 4 or is defined by 4 k+ 5, 4 k+ 7, or 4 k+ 8, where k= 0,1,, this product array will also give the above-mentioned small bounds to the possible unbalanced signal on the output wire. Similarly other modular product arrays could have the pattern 1 00 1 1 00 1 1 ( 7) for the first row, which pattern is again shifted for each succeeding row For cases where N is equal to 4 or is defined by 4 k+ 6 or 4 k+ 8, where k= 0,1,, this product array will also give the desired bounds to the possible unbalanced signal on the output wire. Having chosen a suitable modular product array, one can easily determine the output array, and thus the threading of the output wire, through the matrix, by comparison of the product array with the selection array. This can be done by a visual comparison or mathematically by the

relationship given in equation ( 3) above To illustrate the practice of this invention this will now be done for the specific matrix we have been considering, namely one in which n= 7. As stated above, the selection array S is given by expression ( 1) for any pierced type of matrix and is 101 101 S= O 1 O 101 101 0101 1010 0101 1010 0101 1010 0101 Let us assume that the product array has a 1 1 O O 1 1 O O 1 1 O O ( 9) giving the product array 1 1 00 11 O 1 O 01 10 1 00 11 01 1 P= O 1 1 0 11 0 1 1 01 10 O 1011001 0 1 1001 1 ( 10) By comparison of P, expression ( 10), and S, expression ( 8), we can determine the output array Q, which is given by 1001 100 0011000 0110001 Q= 1 1 O O O 1 1 1000110 0001100 0011001 ( 11) With expression ( 11) as our guide, the actual threading of the output wire through the cores of the matrix becomes simply a question of threading successive cores so as to economize on the length of the output wire between 85 these successive cores; the difficult problem of the direction in which the output wire intersects or threads the core has been resolved by expression ( 11). Fig 3 is one wiring schematic for the out 90 put array Q of expression ( 11) In this embodiment the output or read-out wire 18 is threaded through each core starting, as indicated by the arrow, at the core in the first column of the second row Because for this 95 matrix N is an odd number, it is desirable to thread the output wire through each core along the diagonal of the matrix As the diagonal elements of the output array all have the same value, it is necessary for the read 100 out wire 18 to thread each core in the same direction; accordingly, a set of extra holes 20, indicated as square holes in the figure, extend through the element defining the matrix and are threaded only by the output wire 18 105 Obviously, the holes 20 need not define magnetic cores Except for the diagonal cores and certain cores at the edges of the matrix, successive cores intersected by the read-out wire have opposite values in the output array 110 For this particular array and the particular core selected, as depicted in Fig 1, let us 785,897 6 785,897 consider the summation of the error pulses on the output wire 18 in Fig 3 This can most readily be done by considering the product array P, with those elements underlined that have a disturbing pulse applied thereto and therefore contribute an error pulse to the read-out wire; again the actual core selected appears at X. 1 1 O O 1 1 O 1001101 O -O1 11 P=o Yi o T 1 i O T 1 10 EOXO 101 1 O O 1 01 To 0061 1 in this case there is a complete cancellation of the error pulses; This will not always occur; however, the maximum possible unbalanced signal per set of selection wires is between 0 and + 2 when N is an odd number, as mentioned above, and this is true regardless of the size of the matrix or the number of selection wires. In the above description and example, the product array has been

defined by shifting the modular pattern of the first row one core or element to the left for each succeeding row. The pattern of the first row may, however, be successively shifted, either to the right or to the left, by any of a number of cores; the major requirement is that the amount of the shift be a prime number relative to n If we consider the amount of the shift to be s, then s and N are relatively prime In many cases, however, the simplest wiring arrangement for the read-out wire will be for the case s= 1. Advantageously, each selection wire is threaded through the cores of the -matrix so that no two selection wires jointly thread any two cores, as described in detail in the abovementioned British patent. Further, in order to attain the minimum unbalanced signal, the amount of the shift s should also be related to the particular selection wires utilized in the array Specifically, as described in the above-mentioned British patent, each set of selection wires can be defined by a pair of numbers (ab,) where a is the distance between successive cores threaded by the wire in the a coordinate direction and b is the distance between successive cores threaded by the wire in the b coordinate direction Further, each selection wire, except the ( 0,1) and ( 1,0) wires, which are the coordinate selection wires, threads only one core in each row and column of the matrix Accordingly, we can consider each set of wires other than the coordinate selection wires to be defined by the pair of numbers (tl) where, starting at any core threaded by a selection wire, we determine the distance t in the a coordinate direction to the core in the next row in the b coordinate direction that is also threaded by that selection wire. Thus t is the distance between two cores in the a coordinate direction threaded by the same selection wire which are separated by a distance of unity in the b coordinate direction; it should be pointed out that these two cores need not be threaded successively by the same wire. To obtain a minimum unbalanced error signal, each set of selection wires which is defined by a pair of numbers (t,1) should be chosen such that the value (t-s) is a prime number relative to n It may be noted that if s is chosen as unity, and if N is prime, t may take any value between 0 and n-1 except unity. It is to be understood that the above arrangements have been described merely by way of example and that other output arrays and wiring arrangements may be devised by those skilled in the art without departing from the scope of the invention as defined by the appended claims.

* Sitemap * Accessibility

* Legal notice * Terms of use * Last updated: 08.04.2015 * Worldwide Database * 5.8.23.4; 93p

* GB785898 (A)

Description: GB785898 (A) ? 1957-11-06

Improvements in draught excluding devices for doors or windows

Description of GB785898 (A)

PATENT SPECIFICATION 758,898 , Date of application and filing Complete Specification: Nov4, 1955. No 31590/55. Complete Specification Published: Nov 6, 1957. Index at acceptance:-Class 20 ( 3), D 2, J 2 (H: J). International Classification:-E 04 f. COMPLETE SPECIFICATION Improvements in Draught Excluding Devices' for Doors or Windows 1, JORDAAN KNAAP, a Dutch Subject, of V Westersedriit 97, Haren, Holland, do hereby declare the invention, for which I pray that a patent may be granted to me, and the method by which it is to be performed, to be particularly described in and by the following statement: - This mvention relates to draught excluding devices for doorsl and windows. Resilient weather strips located between a door or window and the door or window frame whereby the gap between the door or window and the frame is effectively sealed against draughts are already known In one known ararngement a thickened portion of the weather strip is enclosed in a hollow tube-like part of a sectional strip, which is connected with either the door or window or the door or window frame; and a resilient lip of the weather strip projects outwards from the said thickened portion and engages resiliently against the door or window frame or the door or window according to whether the weather strip is connected with the door or window or its frame.

The object of the present invention is to provide an improved construction of weather strip arrangement of the above described type which can easily be applied to different forms of door or window construction. According to the invention a draught excluding device of the type above described comprises a resilient weather strip having a thickened portion along one edge held in a part of a sectional strip, adapted to be attached to the door or window or to the door or window frame, wherein the sectional strip comprises a web portion for securing said strip to the door or window or said frame and having along at least one of its longitudinal sides a hollow beading which encloses the thickened portion of the weather strip in such a manner that a resilient or elastic tongue or lip of the weather strip projects outwards and engages the said frame or the door or window respectively when said door lPric_ or window is closed, and wherein said hollow beading or a second beading, forming part of the sectional strip, forms an abutment or stop for the door or window 50 In order that the invention may be more clearly understood various embodiments thereof will now be described with reference to the accompanying drawings wherein, Figures 1 to 6 are cross-sections of the 55 draught excluding device according to the invention as applied to different constructions of doors or windows. In the embodiment shown in Figure 1 the sectional strip comprises a flat web portion 1 60 by which said strip can be attached to a door or window frame 7 in any suitable manner. Extending longitudinally along each side of said web portion is a hollow tube or beadlike portion 2, 3 (hereinafter called "hollow 65 beading") of substantially rectangular crosssection A thickened side 4 of a weather strip is enclosed in said hollow beading 2 and through a longitudinal slot 5 in said hollow beading 2 projects a resilient or flex 70 ible tongue or lip 6 of the weather strip The wall 10 of the hollow beading 2 serves as a stop or abutment for the door or window 9 when the latter is closed, said door or window 9 being hinged to, the frame 7 by means 75 of a hinge 8 In this embodiment, which is constructed especially for doors or windows which open outwardly, the hollow beading 3 is recessed in a channel in the frame 7 as shown An alternative method of arranging 80 the sectional strip for doors or windows which open inwardly is illustrated in Figure 2 In this arrangement the thickened side 4 of the resilient weather strip is enclosed in the hollow beading 3 and the opposite hollow 85 beading 2 serves as a stop or abutment for the door or window Thus this embodiment differs from that shown in Figure 1 only in that the sectional strip is turned through ' so that the position of the hollow bead 90 ings 2 and 3 are reversed and the tongue or lip 6 of the

weather strip projects outwards instead of inwards relative to the frame 7. Figure 3 illustrates the manner in which the arrangement according to Figure 2 functions if the door or window 9 or the frame 7 becomes warped or when said door or window fails to close completely for any other reason. The direction of potential draughts is indicated by arrows in Figures 1 and 2 and it will be seen that by means of the present invention the weather strip per se is arranged in such a manner that the potential draught operates to force the resilient tongue or lip harder against the door or window 9 thereby providing a particularly effective seal. The sectional strip illustrated in Figures 1 to 3 can be applied equally well to sash window and the like as shown in Figure 6. In this case the web 1 of the sectional strip is secured in a recess in the window 11, the resilient tongue or lip of the weather strip being arranged to engage in a recess in a covering member 12 As in the above described embodiments the resilient tongue or lip 6 effectively excludes draught and the hollow beading 2 or 3 or both act as stops or abutments which engage the covering member 12. Fiaures 4 and 5 illustrate the application of the invention to window sills In each case the sectional strip is provided with a hollow beading 2 along only one side of the strip said hollow beading serving both for housing the thickened edge 4 of the weather strip and as a stop or abutment for the window Figure 4 shows the application of the invention to an outwardly opening window and Figure 5 an inwardly opening window, the arrangement according to Figure 5 also being provided with leak shields or frames 13 and 15 The member 13 being provided with a face 14 so arranged as to be in the same plane as the wall 10 of the hollow beading 2 thereby forms an additional stop or abutment for the window.

* Sitemap * Accessibility * Legal notice * Terms of use * Last updated: 08.04.2015 * Worldwide Database * 5.8.23.4; 93p