US2577809A - Cold cathode electric discharge tube - Google Patents

Description

Dec. 11, 1951 v A. H. REEVES ErAL 2,577,309

com) CATHODE ELECTRIC DISCHARGE TUBE Filed Jan. -21, 1949 4 Sheets-Sheet 1 l l l i m i I I J F/GZ.

INVENTORS H156 EVA/Hf Y REEVES GEORGE Ill/BERT HUI/6H A TTOPA/l') Dec. 11, 1951 H. REEvEs ETAL COLD CATHODE ELECTRIC DISCHARGE TUBE 4 Sheets-Sheet 2 Filed Jan. 21, 1949 FIGS.

ATTORNEY Dec, 11, 1951 A. H. REEVES EI'AL COLD CATHODE ELECTRIC DISCHARGE TUBE Filed Jan. 21, 1949 4 Sheet s-Sheet 5 A TTOEWEY Dec. 11, 1951 A. H. REEVES EI'AL 2,577,809

COLD CATHODE ELECTRIC DISCHARGE TUBE Filed Jan. 21, 1949 4 Sheets-Sheet 4 F 6. ll.

F/G. I2.

INVENTORS ATTORNEY Patented Dec. 11, 1951 UNITED STATES PATENT OFFICE COLD CATHODE ELECTRIC DISCHARGE TUBE Application January 21, 1949, Serial No. 72,046 In Great Britain January 22, 1948 16 Claims. I

The present invention relates to cold cathode gas-filled electric discharge tubes of the type in which the striking potential or one gap is reduced by the presence of a discharge at a neighbouring gap in the same envelope.

It is an object of the present invention to provide a cold cathode gas-filled electric discharge tube in which ionisation products may migrate from the first of a pair of glow-discharge gaps to a second discharge gap more readily than from said second gap to said first gap, whereby the striking potential of said second gap may be lowered by discharge at said first gap, but the striking potential at said first gap is substantially independent of the discharge at an anode potential of said second gap.

Although numerous possible applications of such a tube will occur to those skilled in the gas tube circuit art, the present invention has been conceived with the object of enabling two or more sequence discharge arrays to be used in a single tube envelope and to have substantially uni-directional coupling between them. Sequence discharge tubes are disclosed in co-pending application No. 763,655, filed July 25, 1947. by A. H. Reeves, and may be defined for present purposes as cold cathode tubes having an array of gaps so arranged that when one gap is fired, ionisation products migrate into the neighbouring gap and reduce its striking potential. In this way successive voltage pulses applied in common to the electrodes of the array may cause gaps to fire in sequence from a given starting gap. Such tubes may be used, inter alia, as electrical counters in calculating machines or as message registers in automatic telephone circuits. If desired, a steady polarising potential may be applied between the anodes and cathodes of the gaps of the sequence arrays, so that a gap, once fired, may remain in a discharging condition indefinitely. In order to extinguish the discharges, the polarising potential must be removed, or reduced below a critical value-the maintaining voltage-for a time sufiicient for deionisation to take efiect. Thus, if high rates of counting are envisaged it may be necessary to use a pair of tubes alternately, one counting while the other is being deionised. As disclosed in co-pending application No. 31,323, filed June 5, 1948, by A. H. Reeves, two or more sequence discharge arrays may be mounted in a single envelope and arranged so that ionisation coupling between the arrays may enable second counting array to take over" from a first array while the discharges at the first array are extinguished.

It is a further object of the present invention to provide a sequence discharge tube having two cold cathode discharge arrays with unidirectional coupling between corresponding discharge gaps or the two arrays.

It is yet another object of the invention to provide such a tube in which the discharge arrays are so dimensioned that any given number of the successive gaps in the first array may be fired and maintained in a discharging condition for an indefinite time during which a sequence discharge may take place along the other array to provide information as to the number of gaps of the first array that have been fired.

In the prior applications referred to above, no distinction was made as to whether the coupling between gaps was ionic or electronic, all ionisation products being referred to generally as ions. Certain embodiments of the present invention rely on discriminatory treatment of the migration of positive ions and electrons between gaps. Thus in one type of embodiment, by the application of suitable polarising potentials, ions from the first gap are caused to migrate to a second gap, while ions from the said second gap are inhibited from migrating to the first gap. In other types, unidirectional electronic coupling is used, there being negligible ionic coupling.

Embodiments of the invention will be described with reference to the accompanying drawing in which:

Fig. 1 illustrates diagrammatically one particular use of a discharge tube according to the present invention;

Fig. 2 shows diagrammatically two pairs of electrodes in a discharge tube of the type shown in i 1;

Fig. 3 shows an arrangement or a pair of discharge gaps having unidirectional coupling between them;

Fig. 4 shows an alternative arrangement of the electrodes of a unidirectionally coupled pair of discharge gaps;

Fig. 5 shows a practical embodiment of a discharge tube according'to the invention;

Figs. 6 and 7 show further electrode arrangements according to the invention;

Fig. 8 is a diagram for explaining further embodiments of the invention;

Figs. 9 and 10 illustrate still further electrode arrangements according to the invention, and

Figs. 11 and 12 are circuit diagrams to illustrate the use of certain embodiments of the invention.

In Fig. 1, the gas tube I comprises two . discharge arrays 2 and 3, each being shown as having a common anode 4, 4' respectively and four individual cathodes, 5, 5, 6, 6', I, I and 8, 8', the primed numerals referring to array 3, the others to array 2. In addition, a separate cathode 9 is shown; 9 is a priming cathode and forms a priming discharge gap with anode I in order that the cathodes 5,.5' may be the first to fire in the respective arrays. Anode 4' is connected to a constant voltage source H) which is reprecensus 3 sented diagrammatically as a battery in series witharuistance. Anodel'isalso connected to a signal input terminal ll. Cathodes I, 8' and I are shown connected to ground,

through individual current limiting resistances' i2, while cathode I also feeds apulse'output transformer ll. Cathodes i, I, I, l are connected Inpracflcethetwocathodearraysarearranged paralleltooneanotherbelowtheanodesasshown veryinthe end view of Fig.2. The voltage supplied by source II is adjusted sothatitisnotsuificienttoinitiateanydischargcsatgapsofeitherofthedischargearrays, but will mainhin a discharge onceestabiished. Cathode I isarrangedtodischarge continuomly.

Theoperaiionofthecircuitisasfollowma trainofposiflvepulscs'.whichwewillreferto asdialiirz'pulscaisappliedtotheprimaryterminalsoftransformer It. Thue pulses should be limited in amplitude and duration so that. the first puke discharge takes place between cathodelandanodeLthisdischarge-gapbeing partiallyaiergiscdbythepriming discharge at cathodel. Noothergapshouldfireonthisfirst pulse. Due to the voltage-source II, this discharge may now be maintained indefinitely. A secondpulsewillcausecathodeitofire. Athird willinitiatedischargeatcathodelandafourth at cathode I. Atterminal ll there'are applied repeated trains of pulses of general waveform. as shown at is. This waveform comprises'a series of positive pulses followed by a negative pulse. The positive pulses set up discharges at thegapsofarrayi-whichweshallcalithe coimting arraywhile the negative pulse takes anodel'bdowthemaintaining voltage for suiilcient time to extinguish all at this array. Thepositivepulsesmayeachbemodulated in time position or width, and may typically beoneinmeanwidth. Therepetition rate forthepositive pulses atterminal Ii maybeoftheorderof10llkc./s.whilethe diallingpulsesmaybeappliedwtransformeril atarateoflZeyclesper/sec. Weshallassume thateachoi'theposiflvepulsesappliedtothe counting array is amociated with a definite speechchannel. supposenowthattwodialling pulses are applied to transformer It. These shouldcausecathodeslandi'tofireandranain glowing. Now.whentheiirstpulseofpulsetrain ilisappliedtoterminalilbothgapsassociatsd withcathodealfandt'wiliflreifitbearranged thationisatimifmmthediallinggapsprimesthe immediately adiacmt counting gaps. Thus. the secondpubeofthetrainappliedatilwillinitiate adischargeatcathodelandthethirdwillfire cathodel'. hunthedischargeatl'acorrespondingpubwillawearatthesecondaryterminalsdftransfotmm-il. Thefourthandlast positivepuhewillhave-no effect, as allcathodes ofthecotmtingarrayarealreadyglowing. The negative pulaethm extinguishesthedischarges and the sequence is repeated. Thus, in successiv pulse train, the third pulse appears at the output terminals of transformer II. In the same way, had, say, four dialling pulses been applied to the primary of transformer It, the first pulse of train it would have appeared at output cathode 0'. In general. if the tube had 11: gaps in dialling and counting arrays and a dialling pulses were applied to It, pulse No. (n--m+l) would appear on the output cathode. What is more. and this is one of the principal objects of such a circuit, if through interference or other temporary cause, the counting sequence is upset on any one train, it will be corrected for the next train because the selecting meansthe dialling pulses-sets up a permanent selecting condition.

It is to he understood that the circuit shown in Fig. 1 has been introduced merely for general explanatory purposes and in practice alternative circuit arrangements may be provided. Similarly various modifications to the electrode arrangement of tube i are pouible-such as providing a common internal connection for cathodes I. I. l' and I.

We can now consider the practical difiiculties which must be overcome in connection with the discharge tube before a circuit such as has been described may work. Inxthe first case. there is a large difference between the operating speeds required for the two discharge arrays; one is to operate at dialling speed-l2 cycles/see.and the other at a pulse repetition rate of 100 kc./see. As explained in copending' application No. 783,855, gap dimensions and spacings in sequence discharge tubes are largely determined by the counting rates required; on the other hand, in order to provide coupling between corresponding gaps in the two arrays, the gap spacing should be the same for each. In order to avoid what we call -running"sequential firing due to ionisation spread without the additional potential supplied to the gap by signal pulses-the spacings between cathodes in the dialling array must be made as large as posible and the cathode currents as small as possible for a given gas mixture and pulse voltage. Before quoting suitable dimensions for the discharge gaps, it is advisable to consider the other diificulties to be overcome.

Assuming that we have solved the problem of providing a pair of discharge arrays having similar inter-cathode spacings but capable of satisfactory individual operation at widely different counting speeds. we must next consider the question of interaction between the two discharge arrays, when they are operating simultaneously. The first problem is that of priming the gaps of the counting array from the discharges being maintained at the dialling array without discharges at the counting array priming the dialling array and thus increasing the tendency for self-running or an unstable registering condition in the dialling array. Thus, consider the case whenadlallingcountofilhasbeensetupinthe gg dialling array. We require that cathode glow shallbemaintainedoncathodeslandlofarray I. When during the counting pulse train applied to array 8, cathode 1' and I are fired, there is a danger that the discharges here may prime 1 either or both of the last two gaps defined by I and I with model of the array 2 so that diachargu at this array tend to step along with the discharge sequence at array I. Some assistance in this direction may be obtained by biassing the 7:, cathodesofarraylsothattheyarealwayspoaitive with respect to those of array 3. In this way positive ions will tend to be attracted to the cathode of lower potential, while electrons will tend to keep to the direct discharge path from cathode to anode. This in itself, however, will tend to aggravate a further difliculty, that of cross-coupling. Thus, we must prevent discharges occuring between, say, cathode I and anode 4', or between cathode l and anode 4, otherwise selfrunning may also occuron the counting array. Careful consideration must, therefore, be given to these related problems and the present invention resides largely in their solutions Referring .now to Fig. 2, which shows diagrammatically two pairs of electrodes in a discharge tube of thetype described with reference to Fig. 1, let us consider the general double-gap'system there represented. For this purpose anodes 4, 4' and cathodes 5, 6' may represent the electrodes of any pair of gaps having unidirectional coupling properties according to the present invention, but their relative positions as shown are not necessarily such as will result in a practicable tube.

The systems shown in Fig. 2 presents four principal discharge paths, namely the direct paths 3-4 and 5'4' together with the cross-paths 5-4 and 5'-4. Let the maintaining voltages for these paths be V4 5' V4' a" and 5' V4 5 respectively. As a rough guide it may be stated that themaintaining voltage Vm for gaps of difierent length varies according to an equation of the form where k and V are constants and d is the gap length. Since it is required that the gaps 4-5 and 4'-5' are to remain discharging once a discharge has started, external battery voltages may be assumed to provide polarising potentials of V4 and V4 5' across the gaps 4-5 and 4'-5'. In order to ensure that no discharge can be maintained in the gaps 4'5 when cathode 5 is conducting, we must have V4' 5 V4' 5'+P4, assuming the cathodes to be substantially at the same potential where P4, is the pulse voltage applied to anode 4', while to ensure that no discharge may occur under these conditions across the gaps 4-5 we must have V4 5' V4 5. Thus, for a given value of Po, the minimum distance d4 5 is determinedas is also the minimum value of d4 5'.

Taking into account the fact that we wish to bias the dialling cathode positive with respect to the counting cathode in order to make the ionisation coupling from dialling to counting gap greater than in the reverse direction, the above considerations will, in general, lead to different gap lengths for dialling and counting gaps. Another method of achieving the desired relationshipbetween the electrodes is to stagger the anode wires so that gaps 4-4 and 45' are no longer parallel to one another.

In order to assist in making the coupling unidirectional, we have made use of the fact that. during discharge, a large part of the potential fall between anode and cathode occurs in the cathode dark space, which, in the tubes under consideration, extends a very short distance from the cathodeand terminates in the cathode flow. In and around the cathode glow there is an ionic space charge, while the current from the cathode to this space charge region is largely electronic. Thus, if the cathode of the counting gap cannot be "seen by the dialling gap anode, there will be a minimum of electronic coupling in the direc- 6 tion from counting gap to dialling gap, while by making the dialling cathode positive with respect to the counting cathode, ions will tend to migrate from the dialling gap to the counting A preferred arrangement for counting and dialling gaps is shown in Fig. 3, in which the dimensions are considerably exaggerated to show the main features. All electrodes are of nickel, the anodes being-round wires running the length of each array and the cathodes being rods of say, i m. m. diameter coated with alumina except for the actual discharge surfaces 20, 20'. The surface 20' on cathode 5' is inclined so as to be shielded by the top edge of the cathode from anode 4. The anode 4' is arranged to lie on a normal to surface 20', while the two gaps are also staggered in vertical position with respect to one another. For a gas mixture of 92% neon 7% hydrogen and 1% argon at a pressure of m. m. of mercury, and for dialling and counting speeds of 12 cycles/sec, and 100 kc./s. respectively, dialling and counting cathodes are placed 2 m. m. between centres while the gaps 4-5 and 4'-5' are each 1 m. m. long. The cathode-cathode separation along the two arrays may be made 1 m. m. t

A further alternative arrangement is shown in Fig. 4 in which they are so mounted that cathode 5 effectively shields anode 4 from cathode 5' and hence prevents electron coupling from gaps 4' and 5' to gap 4-5.

A practical construction for a discharge tube according to the'presnt invention is shown in Fig. 5. In this tube there are two cathode arrays 2| and 22, each comprising four 1 m. m. diameter nickel rods 23 welded to valve pins 24 in a conventional glass-based envelope 25. The cathode rods are spaced apart by and form supporting means for a pair of mica discs 26 and 21. These, in turn carry three support rods 28, to which two top mica discs 29 and 30 are secured. The top mica discs form support for the two anodes 3| and 32 which are formed of smooth nickel rods bent around three sides of a rectangle and locked in mica discs 29 and 30. Leads from the two anodes are taken to the top caps 33 and 34 respectively. The general arrangement of the electrodes is similar to that shown in Fig. 4. The rods 23 in each cathode array are spaced on 3 m. :11. centres, while the gap length between anode 3| and the cathodes of array 2| is 1 m. m. The gap length for array 22 is 1.5 m. m. Array 2| is spaced 4 m. m. from array 22. Anode 32 is vertically above its cathode array 22 while the anode 3| is displaced slightly from its array 2| and in a direction distally from array 22. The discharge surfaces of rods 23 are ground flat and electrolytically polished, while all other surfaces and internal leads except the anodes 3l, 32 and their connecting leads, are coated with alumina to prevent unwanted discharges. After evacuation through the exhaust tubulation 35, the tube is filled with the neon, hydrogen argon mixture mentioned previously.

In this particular tube, no separate priming cathode is provided, as the tube was designed for three-channel operation, therefore an end cathode of either array 2| or 22 may be used for this purpose.

In the arrangements so far described, the present invention has been put into effect by means of suitable geometrical design of the discharge gaps. It will be appreciated that this involves rather close tolerances both in the tube construction and in the circuit operating conditions. In the embodiments now to be described, novel constructional elements are incorporated in the design which result in simpler construction and wider design tolerance.

As has been pointed out above, a major diillculty in connection with unidirectional couplins between discharge gaps is the phenomenon of cross-firing" between an anode of one main gap and the cathode of another. This cross-firing dimculty may be overcome as in the previously described embodiments, by one or other of the two cathodes acting as a screen. We find that a more direct method is to interpose an insulating screen between the two gaps, leaving a direct path between the two cathodes so that ionisation coupling may take place, this coupling being made unidirectional by biassing one cathode with respect to the other. One such arrangement which has been found to give satisfactory operation in a tube such as described with reference to Fig. l is shown in Fig. 6 in which the gaps l6 and 31 are made of equal length and a sheet of mica II is interposed between the twoanodes l and II. The mica does not extend to the level of the tops of the cathodes 42 and ll. In a satisfactory tube (otherwise similar to that shown in Fig. the mica sheet projected 2 m. m. below the level of the anodes and within 0.5 m. m. of the level of the cathode discharge surfaces; the two cathode arrays were spaced on 2 m. m. centres while the cathode rods of the individual discharge arrays were mounted on 3 m. m. centres. We have found that still better results are obtained, with but little effect upon the ionisation coupling," if the mica sheet is extended between the cathodes and an aperture be left lust sufficient for one cathode discharge surface to be "seen" from the other.v

This arrangement is shown in Fig. 7 in which the electrodes are indicated by the same reference numerals as in Fig. 6 and the insulating sheet 44 has a circular hole 46 in line with the tops of cathodes 42 and 43 and of the same diameter as the discharge surfaces hole I permits migration of positive ions between the two gaps.

As an alternative tothe above arrangements, a very considerable improvement in unidirectional coupling may be obtained by using electronic, as opposed to ionic coupling. To assist in the understanding of the phenomenon involved, consider the effect on the electric field in the region of a glow discharge if a conducting plate be placed parallel to the discharge gap and held at a potential between that of anode and cathode of the gap. Fig. 8 shows roughly the way in which the equipotentials are distorted under static discharge conditions. Anode 40 and cathode 42 are polarised so as to maintain a discharge between them, conducting plate ll being parallel to gap "-42 and being held at a potential below that necesso sary to maintain a discharge from cathode 42. As previously explained, most of the anode-cathode potential fall is localised a very short distance from the cathode surface. Equipotentials 41 are shown in Fig. 8 much further away from 65 the cathode surface than they would in practice, but the diagram does show they are distorted by the neighbouring plate 46; in particular, there is a steep potential gradient near the cathode on the side adjacent plate 46. In consequence, the

discharge path itself is considerably distorted as shown by the dotted line 48. Some electrons from the area of the cathode glow will tend to be drawn off towards plate 46. In particular, if an aperture 4! be drilled through the plate opposite of 8 the steepest potential gradient and an accelerating field is set up on the other side of the plate. electrons will tend to be focussed through the holes and in the direction of the gradient of the accelerating field. For a given thickness of plate 46 and a given potential distribution to the left of the plate, there is a critical value of accelerating field-or rather of the gradient of this fieldto the right of the plate below which the number of electrons which penetrate the aperture I is negligible, and above which a focused stream may be obtained.

Applying this phenomenon to the problem of unidirectional coupling, it isv found that a high degree of unidirectional coupling is obtained by employing an electrode arrangement suchas is shown in Fig. 9, in which anodes ll and ll with corresponding cathodes l2 and II form a pair of discharge gaps as. in the previous examples, but conducting screen or baffle plate 48, placed closer to gap "-42 than ll4l,isolates these gaps from one another except for coupling through the aperture 49. .It may thus be arranged that discharge of gap "42 is virtually independent of the conditions at gap 43, but that when a suiliciently high pulse voltage is applied to anode 4i, the'electrons travelling in the path I. ionise the gap 41-. On the other hand it may also be arranged that discharges may take place at the counting gap "-0 without priming the dialling gap -42.

A sequence discharge tube using an electrode arrangement such as described briefly with reference to Fig. 9 has been used with some success, but fuller description of an t at present preferredwill be given rather than a detailed discussion of the Fig. 9 arrangement.

In Fig. 9, the plate it merely serves as an auxiliary electrode to the discharge gaps 40-42 and 4l4l. It is found, however, that it is possible to dispense with anode II and to make plate 4: function as the dialling anode. The preferred electrode arrangement is illustrated by the cross-sectioned view shown in Fig. 10 and a circuit diagram is given in Fig. 11. In Fig. '10, a

plate so while the counting cathodes u are parallel thereto. Corresponding individual cathodes s: and are mounted so that a plane perpendicular to the plane of the drawing would pass through a cathode 42, an aperture 4. and the corresponding cathode l3 slightly below the upper surface of the latter. On the side facing the dialling cathodes, plate 48 is backed with a mica sheet ll having apertures 52 approximately twice the diameter of apertures ll and exposing an annular surface of plate l0 opposite each dialling cathode. The purpose of mica sheet II will be explained later, its use is not essential to the invention but is preferred when low dialling pulse repetition frequencies are employed.

In the circuit of Fig. 11, the electrodes are denoted by the same reference numerals as in Fig. 10. Each counting cathode 43 is connected to ground through a resistance II, the output cathode also including a pulse transformer (not shown) or other suitable means for passing the "the anode of which is connected to a suitable 9 polarising source 58. The grid of cathode follower 541s supplied with a pulse voltage of waveform similar to that shown at 54, which, however, represents the waveform of the anode 4|. The dialling anode plate 46 is kept at a steady positive potential V: with respect to ground by battery 51. Cathode 42 is connected via resistance 58 to-lead 59 which connects to the remaining dialling cathode resistances 58 (not shown) From lead 59 a connection is made through a stabilised D. C. potential source ill and the secondary of dial pulse input transformer 6| to anode 46. Source 60 serves as a maintainin battery for the dialling array and provides a constant mean voltage V: between anode 46 and lead 59. Dialling pulses are fed in through the primary of transformer 6| from terminals 62. Anode 4| is maintained by the cathode current 01' 54 at a D. C. potential V1 with respect to ground which is increased to V1+P during the counting pulses. The counting pulse train 56 contains a negative extinguis'hingpulse 63 for the purpose of extinguishing discharges at the counting array after each cycle of operation.

In one mode of operation, individual priming discharge gaps (not shown) such as described in connection with Fig. 1 are provided for both dialing. and counting arrays. The pulse amplitude P is then adjusted so that the counting array operates normally without discharges occurring at the dialling array. The potentials V1 and V2 are arranged to be such that, in the absence of counting pulses, no current is drawn from the dialling cathodes when the dialling gaps are discharging. In other words, the potential gradient between 48 and 4| is insufllcient to focus electrons through the apertures 49. This is assisted by having the cathodes 43 a millimetre or so above the level of the respective apertures 49 so that they tend to set up a retarding field. During the counting pulses, however, the in-'- creased potential of anode 4| is arranged to be suflicient to draw current through any 'aperture 49 opposite a discharging dialling cathode. The potential V1-l-P-V2 must then be greater than the anode fall of potential for a gap of length d, and having the geometry shown in Fig. 10, where d is the distance between 4| and the corresponding aperture 49. A stream of electrons then passes between cathode 42 and anode 4| and ionises the gap 4|-43 so that a discharge takes place whether or not the gap is ionised due to discharge at a neighbouring counting gap (or at a priming gap). Thus, at the first pulse of pulse train 56 those counting gaps opposite discharging dialling gaps will fire, the remaining counting gaps, ii any, then firing in normal sequence dischar e manner at successive pulses of the counting train.

At first sight it might appear that the unstruck dialling aps might be fired by the pulse potentials applied to anode 4|. This might conceivahlv occur were a steady potential V1+P to be applied to anode 4|. However. it is to be remembered that in a gaseous discharge, the gap potential difference must be applied for a finite time-the formative delay time-before the gap energy has built u to a value sufilcient to initiate the d scharge. The pulse amplitude P is chosen so that the formative delay time for the gap "-42 is longer than the duration of a counting pulse. The dialling array gaps cannot then be fired by the counting pulses. On the other hand. when a dialling gap is discharging no further energy need be provided by the potential of anode -4| other than that required to draw suiiicient electrons through aperture 49 to ionise the gap 4|-4I; there is then virtually no formative delay time involved in the extension of the discharge (2) In the absence of mica sheet 5| a long-termv there being a less concentrated field across the.

anode cathode gaps of the array and hence a wider diflusion of ionisation products.

The eflect'of (1) is less serious than (2) and can largely be eliminated by the geometrical design of the gaps and the electric field employed: this design is materially assisted by the cure for (2) The long term ionisation coupling (2) manifests itself as a tendency to instability ,of the dialling number registered after a tube has been in operation two or three minutes. The provision of mica sheet 5| (Fig. 10), which 10- calises the discharge to a definite region of the anode, effects a satisfactory cure. Sheet 5| acquires a considerable static charge, particularly at the rim of aperture 52 surrounding aperture 49 which charge reduces the ionisation coupling between dialling cathodes thereby limiting the eflect of the long term ionisation coupling.

In the design to eliminate efi'ect (1) it can be arranged. for example, that the whole of the current at the dialling gap is transferred from the dialling anode to the counting anode. Thus, with a suitable field on the other side of the plate, all the electrons go through the aperture 49 rather than being collected by plate 46. This is illustrated in the following table in which electrode voltages and currents are compared. In the table:

Inc denotes dialling cathode current (electrode 42 in Figs. 10 and 11). 55 Im denotes dialling anode current (electrode 46 in Figs. 10 and 11) Ica denotes countin anode current (electrode 4| in Figs. 10 and 11) VOA denotes counting anode voltage with respect It will be seen that there is an increase 01' only 150 p amp.- in the dialling cathode current for a volt pulse on the counting anode. Since the :5 gap spacings-which are similar to those given incdnnectionwithl'igJ-aresuchthatthereis very diflerentfrom but little variation in ionisation coupling between gaps with discharge currents for dis-- charges-of the order of 400 amp.. the tendency for self-running is largely eliminated.

It was mentioned above, that in one mode of operation, both dialling and counting arrays could be provided with priming discharge gaps to control the general ionisation level in both arrays and so ensure that sequence discharges commenced at the first gap in each' array. The voltages of pulse train It (Fig. 11) were thus arranged so that the counting array counted normally in the absence of discharges at the dialling array. Such a mode of operation is set satisfactorlly when only a small number of gaps-say four-are involved in each array. with arrays capable of dialling and counting higher numbers, some difiiculty is liable to be experienced in the counting array, due to the fact that the general level of ionisation decreases exponentially with distance away from the permanently discharging priming gap at the end of the array. In consequence, due to the phenomenon of formative delay time, when several dialling gaps are discharging. the corruponding counting gaps do not all fire at exactly the same instant on application of the first pulse of train II, it is found thatthereisatendencyforthemstilltofirein sequence with consequent detriment to the exact- Although the discharge tubes described with reference to Figs. 9, l and 11 were devised with a view to'meeting the requirements of a sequence discharge tube as discumed with reference to Fig. l, the phenomena involved in the electronic coupling between-gaps through aperture 0 may be utilised with advantage in a tube other than of the sequence discharge im Thus, Figs. 9 and 10 may be taken to' illustrate electrode arrangements for tubes having single pairs of coupled gaps to utilise the electron coupling phenomenon. Thus. it may be arranged that, if anode ll be maintained at constant D. C. potential and the i2 thoeewith amperes or millamperes which can he embodiments ofthepresmtinventson.

Accordingtothisfmtheramectofthepresent invention, is provided a gas-filled elecil'icdischarge tube comprisingacoldeathode discharge gap between solid electrode surfaces. and a metal screenhavinganapertureseparatingthecathode ofsaidgapfromafurtheranodecrcontrolelectrode,theelectmdesbeingarnngedsothatelectronsfmmthedischargeatmidcathodemaybe drawnthroughsaidapertm'ah-idmrflierelectrodetmdertheinfiuenceofthepotentlalfield betweensaidscreenandmidfnrtherelectrode. ltwillbeevidentthatsuehatnhewbemade tofulfillmanyofthefrmcticnshesetoforeperformed with thermionic valm and that other additionalelectrodesmayhehnertedtoaimulate the various multl-grid tube ol the thermionic valveart.

'Acircuitutilisingathree-eleeisodetubeaccordingtothisfurtheraspectoftherlesantinventionisahowninl'igJz. Inthheirenitanodell isconnectedthroughthehmedaecondarycf transformerlltotheposiflveterminalofbattery whileaperturedplatediisconnectedviaflie primarywindingoflitothenegativepoleofbatteryllmdtothepositivepoleofbatteryllthe negative pole of which is takm through current limitingresistanceiltocathodeu. Anearth conncctioncanbemadewhereindieatedorat potential of plate 40 be raised, the current fiow- M ing to anode ll may be decreased'in spite of the extra current drawn from cathode 42 to electrode ll. Similarly, if electrode 40 be maintained at a steady potential and that of electrode ll be raised, the current to electrode 46 will decrease. It follows that the tube may be made to act as a negative resistance. For many purposes catho'deflcouldbedispensedwithandthetube couldthenberegardedasatriodehavinganegative mutual conductance-electrode ll functioning as anode and-ll as control grid, or vice versa. Such a structure is not dissimilar from those previously described in connection with a tube utilising what is termed a Plasma cathode in which an apertured screen constricts the crosssection of the discharge and enables an impoverishment of positive ions to occur in the anodic space with consequent negative resistance properties between the "Plasma" cathode and the any other convenient point. The operation is as follows:

Batterypotentialsllandllareadiustedto maintain a discharge betwem cathode 42 and platellwithanextensionofthedischargeto anode I. Then if, momentarilmthepotentialof anodellincreaseathecurrentatplateupassingthroughtheprimaryodinmformerlldecreases. Itisarrangedthatthcwindinuof aresopoledthatthedecreaaeinprimarycurrent causesanincreaseinthevoltageodanodell. The potentialacrossthetlmedsecondarybuildsupto alimitingvaluedeterminedbythedreuitconstants and the non-linearity of the volhge-current charaeteristicsofthemafterwhiehitfalls and reverses the directionof regenerative action between electrodes ll and ll. Oscillations are thussetupacmssthehmedseconlhrycireuitof transformer ,andmaybeobtainedfor example throughatertiarywindingilcoupledtotheprimaryandsecondarywindinga.

Thetransformer Clmaybereplacedbyany other suitable type of four-tannin! network or tion ofthecircuitdescribedwithrefa'encetol'ig. I2, aconsiderablyhlshercathodecurrmtisdesirablethanthe400500 ampa.qnotedintheabove table. Increased currentscveraltimesiargermay readilybeobtainedbyincre'asingtheefiecflvearea of cathode 42 without incurringacos'respondingly substantialincrease in the voltages Vc'a and Vna quoted in the table. 'In viewed the complete anode. The operation of such tubes is. however. 1s above in connection with specific ex,

, l3 amples and particular modifications thereof, it is to be clearly understood that this description is made only by wayof example and not as a limita tion on the scope of the invention. a What is claimed is: 1. A cold cathode gas-filled discharge tube comprising a plurality oi electrode arrays defining thereamong at least two discharge gaps, a first of said gaps having a lower given firing potential the cathode of said first array, said portion lying in a plane with and disposed between the anode of said first array and the cathode of said second array.

3. A cold cathode gas-filled electric discharge tube as claimed in claim 1 wherein each of said arrays comprises an anode and a cooperating cathode, each of said cathodes having a dischargeinhibiting area except at the portion adjacent its respective anode, whereby glow discharge is confined to the portion of each of said cathodes adjacent the respective anode, the said portions lying in difierent planes and said cathodes being positioned closer together than said anodes.

4. A discharge tube according to claim 3 in which the discharge portion of said second cathode faces away from the said first gap, the cathode being shaped to shield the cathode glow thereat from the electric field of the said first anode.

5. A discharge tube according to claim 3 in which the said first cathode is positioned to shield the cathode glow at the discharge portion of the said second cathode from the electric field of the said first anode.

6. A cold cathode gas filled electric discharge tube as claimed in claim 1, wherein each of said arrays comprises an anode and a cooperating cathode, the cathode of said first array adapted to be biassed at a higher positive potential than the cathode of said second array and said inhibiting means comprises a bafile positioned to separate said gaps, a portion of said bailie lying in a plane traversed by a line taken between the anode of said first array and the cathode of said second second array, whereby positive ions may migrate from said first gaps to said second gaps.

7. A cold cathode gas-filled electric discharge tube comprising a plurality of electrode arrays each array having an anode and a cooperating cathode and defining a discharge gap therebetween, a first of said gaps having a lower given firing potential than a second of said gaps, the cathode of said first array adapted to be biassed at a higher positive potential than the cathode of said second array, baflle means positioned between said arrays, said baiile, having an aperture therethrough to communicate with each of said gaps, said aperture substantially on a line with the discharge surfaces of said cathodes.

8. A cold cathode gas-filled electric discharge tube as claimed in claim '7, wherein the anode of the array making up said first discharge gap comprises a sheet of conducting material forming a shield adapted to inhibit positive ionic coupling between said first and said second discharge gaps, said sheet having an aperture therethrough to communicate with each of said gaps, said aperture disposed in a plane substantially ofiset from a line taken between said second cathode and said first anode, whereby electrons travel from said first cathode through said aperture to said second anode to prime said second discharge gap.

9. A cold cathode gas-filled electric discharge tube comprising a metal screen anode and a first cooperating cathode defining a first discharge gap therebetween, a second anode and a second cathode defining a second discharge. gap therebetween, said first gap having a lower given firing potential than said second gap, said screen anode disposed between said two cathodes and having an aperture therethrough substantially on a line taken between said cathodes, whereby electrons travel from said first cathode through said aperture to said second anode to prime said second gap.

10. A coldcathode gas-filled electric discharge tube comprising a first metal screen anode having an aperture therethrough, a common cathode disposed on one side of said first anode, a second anode disposed on the other side of said first anode, said cathode, said aperture and said second anode being in substantial alignment, each of said anodes forming a discharge gap with said cathode, one oi' said gaps being through the aperture in said first anode.

11. An electric oscillation generator comprising a cold cathode electric discharge tube having a first conducting screen anode having an aperture therethrough, a common cathode disposed on one side of said anode, a second anode disposed on the other side of said anode, said cathode, said aperture, and said second anode being in substantial alignment, each of said anodes forming a discharge gap with said cathode, one of said gaps being through the aperture in said anode, and a resonant circuit coupled to said second anode.

12. A cold cathode sequence gas-filled electric discharge tube comprising a first electrode array having pairs of mutually spaced electrodes defining a first series 01' discharge gaps, a second electrode array having corresponding pairs of mutually spaced electrodes defin'ingj a second series of discharge gaps, said first series of gaps having a lower given firing potential than said second series of gaps, the cathode of said first array adapted to be biassed at a higher positive potential than the cathodes oi. said second array, means to energy-couple corresponding of said first series of gaps with corresponding of said second series of gaps, whereby firing of any of said first series of gaps lowers the given firing potential of correspondingoi said second series of gaps, and means disposed between said arrays to inhibit energy coupling between said second series of gaps and said first series of gaps, whereby discharges may be maintained across a desired number of consecutive gaps 01 said first array independently of discharge equences occurring across the gaps of said second array while a discharge sequence along said second array commences at that one of its discharge gaps determined by the said number 01' discharging gaps in said first array.

13. A cold cathode gas-filled electric discharge tube comprising two sets of cathodes, arranged in a pair of parallel lines, each set comprising the same number of equally spaced parallel rods mounted opposite the respective rods of the other set; a pair of anodes each cooperating with a respective of said cathode sets, each anode comas'mscc prising a rod extending parallel to said parallel lines defining with its cooperating cathode set a series of discharge gave and constituting anelectrodearramafirstoisaidarrayshavingalower given firing potential than said second array, means to icnically'eouple gaps 01' said first array with gaps 01' said second array, means to inhibit ionic coupling between gaps of said second array withgapsotsaidfirstarraycornprisingabaiiie sheet interposed in parallel between said two arrayasaidsheethavingaphiralityoispaced apertures therethrough, each aperture substantiallyonalinebstweenthedischargesuriaceso! co cathodes oi the two sets.

14. A coldcathode gas-filled electric discharge tubeasclaimedinciaimimwhereinsaidbaflie andaseeondelectrodearrayeachoimutually' spaced electrodes defining discharge gaps adapted tofireinsequenoealongthearraywhenenergized by voltage pulses applied in common to the electzodesoi' that array, the gaps of a first of said arrays having a lower given firing potential than the gaps of said second array, the electrodes in the two arrays being positioned relative to one another to provide successive pairs of discharge gaps, the gaps of said first array being in ionization coupling relation with corresponding gaps of said second array, meam i'or inhibiting ionisation coupling 1Q fromgapsoisaidseeondarraywithcorrespondinggapsoisaidfirstarrayoomprisingportionsoi like eleetrodesin'eachoitheganloisaidfirst array, each of said portions lying in a plane with and disposed between thecooperating electrode 01 each oi said gaps and the like electrode of said second array, means for maintaining discharge acrossfiredgapsotsaidfirstarrayatterthepassage 01' said pulses. means for applying a recurrenttrainotpulsestothesecondarraywher'eby thefirstpulseoieachsaidtrainoi'pulsesinitiates -discharge at the same number of gaps of the secondarrayastherearegapsoithefirstarraybeing maintainedindischar in conditiomtheremainingunfiredgapsotthesecondarraybeingfiredin sequencebysucceedingpulsesoisaidtraima utilization circuit coupled to the last gap of said secondarray;meansi'orpassingasignaltosaid utilisation circuit when said last cap has been fired; and means for extinguishingthe discharges of the fired gaps 01 said second array between each of said'recurrent pulse trains.

ALIC HARLEY REEVES. GEORGE HUBERT ROUGH.

REFERENCES CITED The following references are of record in the file 01 this patent:

UNITED STATES PATENTS Number Name Date 2.415.816 Depew et al Feb. 18, 1947 2.443.407 Wales June 15, 1948 2,473,159 Lyman June 14, 1949

Images