US2520171A - Electric discharge device - Google Patents

Description

ug. 29, 1950 A. H. REEVES 2,520,171

ELECTRIC DISCHARGE DEVICE Filed. April 5, 1948 3 Sheets-Sheet 1 In enlor Attorney 3' Sheets-Sheet 2 Filed April 5, 1948 Attorney 9, 1950 A. H. REEvEs 2,520,171

ELECTRIC DISCHARGE DEVICE Filed April 5, 1948 I5 Sheets-Sheet 3 FIGS.

I nuentor flaw. fwd B Attorney Patented Aug. 29, 1950 ELECTRIC DISCHARGE DEVICE Alec Harley Reeves, London, England, assignor to International Standard Electric Corporation, New York, N. Y., a corporation of Delaware Application April 5, 1948, Serial No. 19,084 In Great Britain May "5, 1947 14 Claims.

The present invention relates to cold cathode electric discharge tubes of the sequence discharge and related types and is particularly concerned with means for adjusting the ionisation level in such tubes.

The meaning of the term sequence discharge type and the other terms used in this specification will become apparent in the following discussion of the general principles of these tubes.

It is characteristic of cold cathode gas filled discharge tubes that the voltage required to initiate a discharge. between two electrodes depends upon the nature and pressure of the gas and its state of ionisation, the shape and material of the discharge electrodes and the distance between themthe gap length. At low interelectrode voltages negligible current will flow if the gas is initially deionised. As the voltage is increased the molecules of the gas becomes ionized until eventually a discharge is set up with a rapid increase in current flow which may typically rise from microamps to milliamps at a critical voltage which is the striking or firing voltage for the discharge gap. The discharge is characterised by a glow which at first appears at the cathode and may extend-to the anode and beyond the immediate neighborhood of the gap, depending upon the degree of ionisation. When once a discharge has been established the interelectrode voltage tends to fall and becomes to a large extent independent of the discharge current. This approximately constant voltage is called the maintaining voltage. Thus in the familiar neon tubes used as voltage regulators and the like, the striking potential may be of the order of 100 volts or more, but the maintaining voltage tends to be constant in the region of 80 volts.

Ions and electrons tend to migrate from the immediate neighbourhood of the discharge. This phenomenon has been extensively used tolower the striking potential of another discharge gap in the same tube envelope. In one known device there is provided a main discharge gap between main anode and a cathode and a trigger gap between an auxiliary anode and the said cathode. The auxiliary anode is much closer to the oathode than the main anode so that the initial striking voltage of the trigger gap is considerbelow that of the main gap. The trigger gap is used to lower the striking voltage of, or to prime the main gap by ionisation coupling.

It is also well known that whereas when striklug voltage is applied to agap, ionisation takes place in a very short time, measured inmicroseconds,.ifthe voltage across a discharging gap be removed, deionisation is not complete =until after'a substantial period of time which may often amount to milliseconds.

: In my copending: application Ser. No. 763,655, filed-July 25, 1947; the phenomenon of ionisation coupling is utilised to provide a cold cathode discharge tube having-anordered array of gaps in Whichthe"starting gap is arranged to-have-a lowered striking voltage than the remainder so that arecurrent voltage applied to each gap in turn or to all gaps togetherfires first the starting gap and .then the remainder in order. Due to the finite 'deionising time, it isnot necessary that apriming gap be discharging simultaneously with the application of the recurrent voltage to a neighbouring gap. Thussuch a sequencedischarge tube, as it may be called, may be'operatedby atrain of pulses applied between a comnion anode in the form of a plate or-wire and'a'n array of cathodesyeach mounted independently, or forming the hills of a corrugated plate facing the anode, or, perhaps, taking the form of rods mounted upon a common plate or wire like the teeth of a comb, and having the gap length'for a given gap, usually the first, shorter than the remainder. 'Then the first or starting gap may be fired by the first pulse, while the second pulse reignites the starting gap and the next adjacent gap simultaneously. The third pulse will fire the first three gaps and so on. On the other hand, if a maintaining battery be connected across the gaps, once fired each gap will remain discharging indefinitely. Thus such a sequence discharge tube has been used, with suitable circuits to extract the information and to restore the initial conditions as to ionisation, as a counter or/and as an information storage device.

Applicant has found with highsionisation coupling that the striking voltage of a gap may be reduced to or even below the maintaining voltage so that ionisation spread irom a discharge proceeds at a definite rate, depending uponhthe conditions in the discharge tube. These phenomone have been utilised to provide sequence discharge tubes in which, after discharge-has-been established at a starting gap, the other gaps of an array may be made to fire automatically in order at definite time intervals, thus-providing means for generating pulse trains. I Sequence Edischargetubes utilising the time aspect of ionisation migration form the subject matter of my copending application'bearing Serial No.- 14 ,184, filed March 11, 1948.

We have found that the rate of spread of ionisation from a discharging gap depends, inter alia, upon the discharge current passed. Thus the amount by which the striking potential of a neighbouring gap is reduced, and the rate of such reduction, depends in part upon the priming discharge current. In the past it has been. proposed to stabilise the operation of electric discharge tubes for illumination purposes by regulating the general level of ionisation in the tube, by means of an auxiliary discharge which is maintained continuously. The present invention is primarily concerned with the use of such an auxiliary discharge to regulate the ionisation level in sequence discharge and related tubes, and thus to alter in some cases their mode of operation, or again to effect the time constants of the discharges.

According to one a pect, the invention provides a cold cathode gas filled electric discharge tube comprising a plurality of main discharge gaps forming an ordered array, and one or more auxiliary discharge gaps placed in relation to one or more of the main gaps in such manner as to permit regulation of the ionisation level at the said one or more of the main gaps by the maintenance of a discharge at the uxiliary ap or gaps.

According to another aspect, the invention provides a cold cathode gas fil ed electric discharge tube compri ing one or more main discharge gaps and a seouence of auxiliary discharge gaps so placed in relation to the main gap or ga'p as to permit a variable control of the ioni ation Fig. 7 shows the electrode arrangement of the device of Fig. 6.

Figs. 8 and 9 show explanatory curves in connection with the circuit of Fig. 6.

It has been explained above that in the gap arrays of sequence discharge devices it is usual to provide a starting gap by making this gap of shorter length than the remainder of the gaps of the array. In Fig. 1 a device I is shown in which the starting gap may be of the same length as the others. The device I comprises an anode 2 in the form of a smooth plate or rod and a cathode formed as a wire helix. Except for a narrow strip facing the anode, the helix is coated with material, such as alumina, which inhibits the spread of cathode glow around the turns of the wire and confines the glow to the uncoated portions. Thus each turn of the helix forms a. separate discharge gap with anode 2. The right hand end of the helix terminates in a length of smooth rod 4 which forms an extinguishing squeg tab as described in my copending application bearing Serial No. 19,083, filed April 5, 1948, and whose function is explained below. Near the other end of the helix an independent cathode 5 is provided to form with anode 2 a primin gap 6. Cathode 3 is connected to ground through resistance T shunted by condenser 8. Cathode 5 p is connected via variable resistance 9 to the negalevel at the main discharge ap or gaps in accordance with the number of auxiliary aps which are discharging.

According to still another asp ct the invention provides a cold cathode gas filled electric discharge tube comprising electrodes forming an extended discharge gap along which a discharge is adapted to spread, and one or more auxiliary dischar e ga s so laced in relation to the said electrodes as to permit regulation of the ionisation level in the neighbourhood of the said electrodes by the maintenance of a discharge across the auxiliary discharge gap or gaps.

The invention also covers circuit arran ements which include means for producing ionisation control, disc arges across the auxiliary discharge ga s of such tubes.

For brevit we shall refer to these auxiliary gaps and discharges as priming gaps or priming discharges.

The nature of the invention as regards the above aspect and others to be discussed later will be more easily understood from the following description of embodiments of the present invention Fig. 3 shows an alternative arrangement to that of Fig. 2.

Fig, 4 shows another type of device and circuit utilising the invention.

Fig. 5 shows explanatory curves relating to the circuit of Fig. 4.

Fig. 6 shows a further type of device and Gilcuit utilising the invention.

tive pole of battery I 0, whose positive pole is connected to ground. Anode 2 is taken via the primary winding of pulse transformer H to the positive pole of battery l2 whose negative pole is connected to ground.

The voltage of battery I2 is insufficient to initiate a dis-charge from any cathode but is suflicient to maintain any or all discharges from cathode 3. Battery l0 assists battery l2 in initiating and maintaining permanently a priming discharge at gap 6, the discharge current being regulated by resistance 9. Assuming that the gas in device I is in a deionized condition and that battery I 0 is put into circuit a continuous discharge will commonce at gap 6. With suitable spacings, gas pressure and other conditions as described in the above-mentioned application bearing Serial No. 14,184, after a definite time the striking potential for the first gap formed by cathode 3 will be reduced to the voltage of battery I 2 and a discharge will be set up across to anode 2. After a further time interval the next gap formed by the turns of the helix will fire, and then the next in order from left to right. The total current through the primary of transformer II will then increase in steps separated by definite time intervals. Hence a train of voltage pulses will appear at the secondary terminals of transformer ll. When the cathode glow reaches the smooth rod 4, the total discharge current rises rapidly but the resistance I limits the discharge current to a value insuflicient to produce enough ionisation to maintain the increased discharge, Hence all discharges from cathode 3 will be extinguished and de-ionisation will set in until the charge on condenser 8 has leaked away sufficiently through resistance 1, when the sequence of discharge will be recommenced from the left hand end of cathode 3 which is nearest to the ion source 6.

It will be seen that the priming gap 5 ensured that the left hand turn of the helix 3 functions as a starting gap. Furthermore the variation in the amount of current passed by the priming discharge when the resistance 9 is varied will affect the rate of spread of ionisation throughput the tubt- In the Position shown in Fig. l,

however, the primin discharge has a greater effect at the left hand end of the cathode arrays. In certain types of sequence discharge tube it may be advantageous to provide each individual main gap with an auxiliary priming gap to control the general ionisation level throughout the tube. Alternatively the main gap array may be arranged symmetrically about the priming gap.

An increase in the general ionisation level in a sequence discharge tube has four principal effects on the gaps of the main array:

1. A reduction of the minimum striking voltage.

2. A reduction of the difference between the striking and maintaining voltages;

3. A reduction in the minimum extinguishing time; and

4. An increase in the rate of spread of cathode glow along each cathode when the discharge is struck.

Effect No. 2 is very useful when it is desired to reduce the interval between pulse trains in a pulse train generator such as described above, while effect No. 4 may be utilised, for example, in a pulse time modulated signalling system to determine the relation between signal voltage and the resulting time modulation, and can be used to adjust this relation.

As an example of this aspect of the invention the case, illustrated in Fig. 2 may be considered in which device 5 comprising a circular anode plate 2, and an array of separate cathodes 3 formed by rods mounted normally to the plane of anode plate 2 symmetrically aboutan auxiliary cathode rod 5. All the cathodes 3 except that indicated 3 form gaps of equal length with anode 2. Sufficient distance is provided between cathodes 3 and 3" to ensure that the discharge sequence shall proceed clockwise around the tube with 3" being the last cathode to fire. Each cathode rod is connected to ground through a limiting resistance and the secondary of a speech transformer associated with a respective separate speech channel. To avoid complicating the drawing, only one cathode 3" is shown connected to its limiting resistance l3 and transformer 14. Connections to the auxiliary cathode 5 and anode 2 arev similar to those previously described in connection with Fig. 1, except that a further connection is made from the anode to terminal l5 through blocking condenser [6.

A master pulse generator (not shown) is connected to terminal l5 and supplies positive starting pulses and negative extinguishing pulses tothe anode. The starting pulse is arranged to fire thE.TfiI'Sii gap from cathode 3. Assuming that no speech currents are flowing in any of thetransformers M, a train of equally timespaced pulses will be obtained from the secondary oftransformer ll. After the last cathode 3" hasfired, an extinguishing pulse is applied at terminal l5 and quenches' all the discharges except'that from the priming cathode 5. In the presence of speech currents in the primary of a transformer l4, however, the potential of the associated cathode varies in accordance with the speech, so that the time of firing is retarded or. advanced. Furthermore, the unmodulated positionof each pulse will depend upon the time positions of all the preceding pulses in the train and therefore depends upon the modulation of the previous channel pulses. In consequence the output of transformer ll will consist of trains of pulse-period modulated pulses of the type disclcsed'in copending'application Ser. No. 756,262,

6 filed June 21, 1947. The depth of modulation of the pulse may be adjusted by control of the priming discharge current by means of variable resistance 9.

As an alternative to the device of Fig. 2, the device illustrated diagrammatically in Fig. 3 may be used. The arrangement is similar to that of Fig. 2 except that an auxiliary anode ll, in the form of a disc insulated from anode 2, which is now annular, co-operates with cathode 5. The battery 10 of Fig. 2 may be omitted.

As applied to a discharge device comprising a pair of smooth opposed electrodes forming the main discharge gaps, an embodiment of the invention is illustrated in Fig. 4 which is concerned with a device and circuit for de-modulating a pulse modulatedsignal and forming part of the subject matter of my copending application bearing Serial No. 14,185, filed March 11, 1948.

In a multichannel elcctric pulse communication system, the usual method of demodulating pulses which are signal modulated either as to amplitude, duration, or time of occurrence, is to pass the pulses of a given channel, or pulses derived from them, through a low pass filter. As the pulse durations are usually very small compared with the repetition period, this method has the serious disadvantage of giving an output power very small compared with the pulse input peak power (unless a valve amplifier is used, which is often to be avoided). The circuit shown in Fig. 4 gives an output power comparable with the peak input power, without hot-cathode valve amplifiers. For a channel system, this may give an output power gain of up to 10,900 over the usual system.

In Fig. 4 the circuit comprises a tube l containing a gas, or suitable mixture of gases, at a suitable pressure. 2 is an anode, which may be, for example, a rod, fiat strip or flat plate, connected to the positive terminal of battery l2. 3 is a cathode which may also be a rod, strip or plate, connected to the negative terminal of battery l2 through resistance ll and one winding of transformer l8. 3 is set at an angle a. with respect to the anode 2. 5 is an extra ionisation control cathode, at a suitable distance from 2, and connected to the negative terminal of l2 through variable resistance 9 to the negative terminal of battery la the positive terminal of which is connected to the negative terminal of battery l2. This auxiliary battery Ill provides an increased potential for the gap between 2 and 5, so that a permanent discharge is produced in this gap.

The pulses to be demodulated are applied between terminals 'l5 and I5 through blocking condenser l6 and decoupling resistor E5, to the cathode 3 in such a sense as to make 3 more negative with respectto 2. The potential of battery [2, the nature and pressure of the gas, the nature of the electrode surfaces, and the smallest gap at end 3 between cathode 3 and anode 2, are arranged so that in the absence of pulses at I5, a glow discharge is not struck at the end 3' to the anode 2 but the pulse voltage at I5 is adjusted so that when the pulse arrives, a glow discharge does occur between 3 and 2. This discharge will first strike at end 3' of 3.

For the duration of the'pulse, the will spread rapidly from the end 3' along'the surface of 3 from left to right, at a speed 'de pending chiefly'on the gas constants, the sum of the pulse voltage and the voltage of battery ionisation I2, resistance I1, angle a, the degree of roughness of cathode surface 3, the control current between cathode and 2, and the spacing between 5 and 3. These various constants are adjusted so that a modulated pulse of maximum duration or of maximum amplitude just enables the glow to spread over the whole length of cathode 3 before the pulse disappears.

If then the modulated pulse has less than this maximum value of amplitude or duration, the glow will not have had time to cover the surface of 3 completely on disappearance of this pulse: experiment and theory both show that the rate of spread is definite quantity. Theory and experiment also show that on disappearance of the pulse the glow may be made to collapse again, and to remain only at the end 3'; and furthermore the rate of collapse may be made slower than the rate of spread during the period of the applied pulse. The collapse rate may generally be made such that for a pulse of maximum amplitude or duration the collapse is only just completed when the following pulse arrives. The principles involved in a device of this type are discussed in my copending application bearing Serial No. 15,582, filed March 18, 1948. It will be understood that the plate 3 may be corrugated or serrated to provide an array of separate gaps of increasing length.

The process is illustrated in the curves of'Fig. 5. Curve a shows three successive input pulses A, B and C of different duration and curve I) shows the corresponding variations of the current iiowing to the cathode 3.

As shown by curve I) the cathode current corresponding to input pulse A rises rapidly from zero during the period of the pulse, and then falls slowly to a small steady value at D, continuing at the value of D until the arrival of the second pulse B of Fig. 5, a. During the period of the second pulse the current through 3 rises again to a higher value E as the input pulse is longer. It then falls again but to the value D, before the third input pulse C arrives. This third pulse causes the cathode current to rise again rapidly to a smaller value, as this third pulse is shorter than the first. The current then collapses gradually again, the rapid ri e and slow collapse occur indefinitely long as the input pulses occur.

The current from cathode 3 flows through transformer 18, of which the secondary winding is connected through a low-pass filter 20 to the output terminals 2|. If 20 is arranged to pass the modulating signal frequencies but not the pulse frequencies, terminals 2i will receive the signal frequency components of the fluctuations of the total area of the curves of Fig. 5, b. The average area may for example, correspond to an average power of the order of r g of the peak pulse input power. In the usual demodulation system, in the case of 100 channels, the average output power, about which the signal fluctuations take place, will be only about /4,o00 of the peak input power. The present system thus has a relative power gain in this case of the order of 2,500.

In the example just described the priming discharge not only stabilises the operation of the tube, but acts as a variable volume control, determining the relation between the maximiun input pulse width and the peak output power. The priming discharge has been used to control inter alia the amplitude of derived pulses. In this and the preceding example the priming discharge has been continuous. There is no reason, however, why the priming discharge may not itself consist of recurrent pulse trains, and instead of one priming gap an ordered array of gaps permitting a sequence discharge may be used to control the mean ionisation level in a discharge device.

As an example of this aspect of the invention, the circuit of Fig. 6 illustrates a method of obtaining a derived train of pulses, corresponding in time relationship with an applied train but with each derived pulse delayed by an adjustable time period from its corresponding input pulse. The device of Fig. 6 comprises an anode plate 2 and a main cathode 3, in a form of a rod mounted normally to the plane of the anode. Cathode 3 is joined to an extinguishing cathode 4 which is a flat plate mounted parallel to the anode 2 but closer to it than the tip of cathode 3. Mounted symmetrically with respect to the line joining the projection of 3 and 4 on the anode but nearer 3 than 4, are two auxiliary cathode arrays 5a, 5b forming priming cathodes. Each of 5a and 51) comprises a comblike or serrated electrode with all projections except an end one 5 on 5d forming gaps with 2 of the same length as formed by cathode rod 3. The arrangement is indicated more clearly in Fig. 7, in which electrode 4 has been omitted for clarity. The cathode arrays 5a and 5b are formed with extinguishing squeg tabs 4a and do such as described in my above-mentioned copending application bearing Serial No. 19,083 comprises smooth elongations of the array electrodes at their mutually remote ends. The priming cathodes are taken to ground through respective resistances la and lb, a variable resistance 9 and battery I0 and also through the respective con-- densers 8a and 8b. Anode 2 is connected to input pulse terminal [5 via blocking condenser l6 and also via resistance I"! to the positive terminal of battery 12 whose negative terminal is connected to ground. Cathodes 3 and 4 are taken to ground through resistance 1 and condenser 8 in parallel. The primary winding of an output pulse transformer l l in series with rectifier 22, resistance 23 and biassing battery 24 is connected across resistance l. Condenser 8, which need only be a few micro-microfarads in capacity, is not large enough to shunt the output transformer circuit to any appreciable extent. The function of the components 22, 23 and 24 is to pass only the peaks of pulses to the output transformer H. The voltages of batteries l0 and I2 are adjusted so that, while a discharge may be maintained from the oathodes 3 and 4 or from the projections of 5a and 5b, no discharge may be initiated except on application of a positive pulse to terminal l5. Pulses to be applied to' l5 are limited so that the joint voltage of both batteries plus input pulse may initiate a discharge from starting cathode 5, but from no other when'the tube I is in a deionised state.

When a pulse is applied at terminal IS, a discharge will be initiated at 5 and glow will spread from point to point of 5a to a number of cathode projections depending upon the amplitude and duration of the applied pulse and also upon this priming discharge current as regulated by battery i0 and resistor 9. After the pulse has passed, the discharge glow will remain stationary on the points of So until a second pulse arrives. The next pulse will continue the discharge sequence further along 5a so that extinguishing tab 4a discharges. Due to the resistances la and 9, andbattery, ill, a squeg occurs and extinguishes. all discharges in the manner discussed in my abovementioned copending application bearing Serial No.-19,083. At the time, however, that the second pulse arrives at 15, ionisation migration from the neighbourhood of (:Szrwill have primed the nearest gap associated with b; hence this second pulse will initiate a sequence of discharges on 52) which will be similar in character tothat described above for 5a. The pulse trains involved are depicted in Fig. 8, in which curve a shows the pulses applied at terminal 15, curves in and c respectively the discharge currents from 5a and 5b with the discontinuous step by step rise of cur:- rent smoothed out, but the extinguishingpulse shown at A. It will. be appreciated that the mean discharge current through 5a and to may be regulated by resistance Sand hence the general ionisation level in the tubeand particularly in the neighbourhood of cathodes Sand l. The discharge arrays of 5a and 5b thus furnish priming arrays for the gaps at 3 and 1.

Because of the ionisation coupling from-the priming arrays, when the second pulse is applied to terminal l5, cathode 3 will fire, and it will remain discharging after the pulse has passed. Depending upon the amplitude and duration of the applied pulse and upon the general ionisation level in the tube, cathode glow will travel along rod 3 more or less rapidly as explained in for example, my above-mentioned application bearing Serial No. 15,582. Theglow is prevented from travelling along the wire connecting 3 and l by coating itwith alumina, or by making the connections externally of the tube envelope. After a further time, but before the arrival of the third pulse, extinguishing cathode 4 will start to discharge. As its projected surface area on anode 2 is so much greater than that of the rod 3, the cathode current will rise suddenly and set up a squeg due to the resistance 7 thus extinguishin the discharge from both cathodes. On arrival of the next pulse the sequence will be repeated. The curve of current passed by 3 and :2 will be similar to curve at of Fig. 8.

Due to the biassing circuit, 22, 23, and 24, only the peaks B will be passed through the primary oi transformer II, as shown .by curve e whilst if it will be a diiferentiating transformer, the secondary voltage will consist of short pulses as shown by curve I of Fig. 8.

There will appear at the output terminals of transformer i l a series of pulses having similar time intervals between them as those of the input train applied to terminal IE, but delayed by a constant time interval which depends upon the priming discharge current. Thus in effect we have here a pulse delay network the delay period or" which may be varied by changing the priming discharge current.

It should be pointed out in connection with the above circuit, that it may be found necessary, due to the somewhat large current variations associated with the discharges to provide voltage stabilising circuits, for example, of the kind described in my copending application bearing Serial No. 19,086, filed April 4, 1948, now abandoned, which in order to simplify the drawing and circuit explanation, have not been shown.

In the foregoing discussion of the circuit of Fig. 6 we have directed attention to the time interval between the application of a pulse at terminal [5 and the onset of a squeg discharge at cathode .4. and, have considered the circuit as a pulse time delay network. If we consider the time intervals between the commencement of discharge from cathode 3 and the squeg discharge of cathode 4, a further aspect of the priming discharge ionisation level control-and another. use for the tube of Fig. 6 will become apparent. If, as stated above, cathode 3 is in the form of a rod, and if the cylindrical surface of the rod be made smooth, then it can be considered that cathode 3 has a smooth surface inclined at an angle of degrees to the anode plate 2. In my above-mentioned copending application bearing, Serial. No. 15,582 we have pointed out that ior smooth opposed electrodes,

cathode glow will spread along the surfaceof the cathode at a rate depending, inter alia, upon the applied voltage. Now the priming of the adjacent squeg gap (cathode fi-anode 2) will depend to a large extent upon the amount of cathode glow on cathode 3 as well as upon the ionisation level due to the priming discharge at ea and 5?). Hence if a speech transformer be inserted between resistance l and ground, speech voltages will modify the glow spread on 3 and hence the time interval between the commencement of discharges on the. two cathodes 5i and 4. Furthermore the law relating applied speech voltage and resulting time modulation of the output pulses may be controlled by adjusting the inclination and surface shape of cathode 3to anode 2, so that we shall now remove the restriction that they be mutually perpendicular. Assuming a linear modulation law, two curves illustrating the relationship between total cathode-anode voltage a and time interval 15 between firing of cathodes 3 and 4 is shown in Fig. 9. The two curves refer to two different states of ionisation as determined by the priming discharges of 5a and 5b. Let the momentary speech voltage bring the voltage 12 to some ordinate between F and G. Then on curve AB the time interval t will vary between H and K, while on curve CD, corresponding to a smaller priming current at 5a and 5b, the same variation in v produces a larger, but still linear variation of t. Hence device I may be used as a modulator with a gain or modulation factor control provided by the priming discharge current 5a and 527.

A further modification of the circuit of Fig. 6 and'd'escribed fully in my copending application Ser. No. 24,953, filed May 4, 1948, now abandoned, may be made by rectifying a part of the modulating speech voltages andapplying the smoothed voltage in series with battery l0or even replacing. The latter the gain control is thus automatic. As described above, cathodes 5a and 5b are serrated and parallel to anode 2. The gain control thus proceeds in steps which may be made to follow any desired law. It will be evident that many modifications may be made inthe arrangement and shape of the priming arrays, thus in some circumstances it may be convenient to incline them to the anode 2, or even to make them smooth. If the latter case, however, the glow would tend to collapse down to the shortest gap after passage of the input pulses at terminal l5.

I claim:

1. A cold cathode gas filled electric discharge tube comprising a plurality of main electrodes defining at least one main gap and a plurality of auxiliary electrodes spaced from said main electrodes and defining a source of sequentially firing auxiliary discharge gaps positioned with respect 11 to said main gap to permit a variable control of the ionisation level at said main discharge gap in accordance with the number of said auxiliary gaps which are discharging.

2. A cold cathode gas filled electric discharge tube comprising a plurality of main electrodes defining a plurality of main discharge gaps and a plurality of auxiliary electrodes spaced from said main electrodes and defining a series of sequential firing auxiliary discharge gaps positioned with respect to said main gaps to permit a variable control of the ionisation .level in the neighborhood of at least one of said main gaps in accordance with the number of said auxiliary gaps which are discharging.

3. A cold cathode gas filled electric discharge tube comprising a plurality of cathodes arranged in a circle, a plate anode co-operating with the said cathodes to form a sequence of discharge gaps, and ionisation level control electrode means spaced from said anode and said cathodes and defining an auxiliary discharge gap placed at the center of the circle for controlling the ionisation level in the neighborhood of said cathodes by the maintenance of a discharge across the auxiliary gap.

4. A discharge tube according to claim 3 in which the auxiliary discharge gap is constituted by an additional cathode co-operating with the said anode.

5. A discharge tube according to claim 3 in which the said anode is annular in form, and in which the auxiliary discharge gap is constituted by an additional cathode co-operating with an additional circular anode arranged in the circular space at the centre of the annular anode.

6. A cold cathode gas filled electric discharge tube comprising two electrodes inclined to one another to form a tapered discharge gap, and an ionisation control electrode co-operating with one of the first mentioned electrodes to form an auxiliary discharge gap controlling the ionisation level in the neighborhood of said tapered discharge gap by the maintenance of a discharge across said auxiliary gap.

7. A discharge tube according to claim 6 in which the said further electrode is arranged adjacent to the smaller end of the tapered discharge gap.

8. A discharge tube according to claim 6 in which one of the first mentioned electrodes is corrugated or serrated.

9. A discharge tube according to claim'fi in which one of the main cathodes comprises a rod arranged perpendicularly to the anode, and the other comprises a plate arranged parallel thereto, the two cathodes being connected together.

10. A discharge tube according to claim 6 in which each array of auxiliary cathodes has a squeg tab connected at one end thereof.

11. A cold cathode electric discharge tube comprising two main cathodes co-operating with an anode to form two main discharge gaps having an ionisation coupling therebetween, and two similar arrays of auxiliary cathodes co-operating with said anode and symmetrically placed with respect to said main cathodes for controlling the ionisation level in the neighborhood of the main discharge gaps in accordance with the number of auxiliary cathodes which may be discharging.

12. An electric circuit arrangement comprising a cold cathode gas filled electric discharge tube including an anode and a main cathode inclined to one another to define a tapered discharge gap, an ionisation level control cathode positioned adjacent the small end of said tapered gap and spaced from said anode to define an auxiliary discharge gap, a source of potential connected across said anode and main cathode, and voltage means coupled to said ionisation control cathode for establishing a permanent discharge between said anode and said last mentioned cathode to control the level of ionisation in the neighborhood of said tapered gap.

13. An electric circuit arrangement comprising a cold cathode gas filled electric discharge tube including a plurality of electrodes defining two main discharge gaps, and ionisation level control electrode means defining a sequence of auxiliary discharge gaps to control the ionisation level in the neighborhood of at least one of said main discharge gaps, and first voltage applying means connected to said plurality of electrodes for applying a maintaining potential across said main gaps of insufficient magnitude to initiate a discharge there-across, and second voltage applying means coupled to said level control means for establishing sequential discharges across said auxiliary gaps effective to fire said main gap after a predetermined delay.

14. An arrangement according to claim 13 further comprising means connected to said second voltage means for adjusting the auxiliary discharge current for the purpose of varying said delay.

ALEC HARLEY REEVES.

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

UNITED STATES PATENTS Number Name Date 1,242,014 May Oct. 2, 1917 1,898,626 Healy Feb. 21, 1933 2,007,932 Ruben July 9, 1935 2,443,407 Wales June 15, 1948 FOREIGN PATENTS Number Country Date 644,402 France June 5, 1923

Images