US2917624A

US2917624A - Switching device for magnetic field tubes

Info

Publication number: US2917624A
Application number: US283396A
Authority: US
Inventors: Sternbeck Olaf
Original assignee: Telefonaktiebolaget LM Ericsson AB
Current assignee: Telefonaktiebolaget LM Ericsson AB
Priority date: 1951-08-15
Filing date: 1952-04-21
Publication date: 1959-12-15
Anticipated expiration: 1976-12-15
Also published as: BE513485A; FR1064766A

Description

Dec. 15, 1959 o. sTERNBEcK SWITCHING DEVICE FOR MAGNETIC FIELD TUBES Filed April 21. 1952 3 Sheets-Sheet l 'MPM J d 1 l i INQENTOR OLBF STERNBECK 31g/ Dec. 15, 1959 o. sTERNBEcK SWITCHING DEVICE FOR MAGNETIC FIELD TUBES Filed April 2l, 1952 3 Sheets-Sheet 2 INVENTOR OIJF STRNBECK Dec. 15, 1959 0. STERNBECK SWITCHING DEVICE FOR MAGNETIC FIELD TUBES Filed April 21, 1952 3 Sheets-Sheet 5 INVENTOR ma@ EGA; B W W fo E m T A s MQ L SWlTCllHNG DEVlCE FOR MAGNETIC BEELD TUBES Olaf Steinbeck, Alvsjo, Sweden, assigner to Telefonaktiebolaget L M Ericsson, Stockholm, Sweden, a company of Sweden Application April 21, i952, Serial No. 283,396 Claims priority, application Sweden August 15, 1951 3 Claims. (Cl. Z50-27) The present invention relates to a switching device for vacuum tubes of the kind, wherein an electron beam, emitted from the cathode, can be guided towards a number of anodes or receiving electrodes by means of electric and/ or magnetic fields.

Such tubes of various designs are previously known or proposed. Thus, in the U.S. Patent No. 2,513,260, different designs of the so called trochotron are described, consisting of a high vacuum tube with a discharge chamber, penetrated by a magnetic iield and containing an electron source with an accelerating anode for generating an electron beam, which is bent in the magnetic field and at the same time given a translational movement through the influence of an electric eld and/or inhomogeneities in said magnetic field. The electron beam will thereby re ceive a cycloidal or a trochoidal form and can be selectively directed towards diierent points Within the vacuum tube by means of special controlling devices (control electrodes or spades) in order, for instance, to influence an electrode there (receiving electrode or plate). A trochotron usually contains a number of control electrodes and a number ol receiving electrodes alternating with each other in such a way that a box containing a receiving electrode is formed between adjacent control electrodes. The trochotrons previously described are composed of oblong tubes, having the abovementioned boxes mainly located along a straight line, running from the electron source to a receiving electrode, placed op posite said source.

Another tube to which a switching device (circuit) according to the invention is applicable is the so called coaxial tube, in which a cathode, centrally located within the tube envelope, is surrounded by a number of anodes, whereby the annular discharge space, formed between the cathode and the anodes, is penetrated by a magnetic iield. The electrodes surrounding the cathode in one design of this tube are of two different types, viz. receiving electrodes, arranged to receive the electron beam emanating from the cathode, and control electrodes for guiding the electron beam towards one of the receiving electrodes. The main direction of the magnetic eld in such a tube is approximately parallel with the longitudinal axis of the cylindrical cathode and the receiving electrodes are shaped like rails, parallel to the cathode. Receiving electrodes and control electrodes alternate with each other, the latter electrodes having greater radial extension towards the cathode and having the part facing the cathode placed closer to the cathode than the corresponding parts of the receiving electrodes. Similar to the trochotron, boxes will thereby be formed between successive control electrodes, containing one or several receiving electrodes.

ln one embodiment of the coaxial tube the above-mentioned receiving electrodes are replaced by a common receiving-electrode cylinder, coaxially surrounding the cathode and the control electrodes. In this case, separate receiving boxes between adjacent control electrodes will, as previously noted, be developed, to which the electron beam can be guided by the aid of the control electrodes,

States Patent O whereby the receiving electrode, belonging to thebox, is electrically the same for all the boxes within the tube.

The reason for this design is, that the common receiving electrode has proved to be suitable for certain purposes, for instance for counting electrical pulses, when the diterent receiving electrodes can be directly connected without any intermediary impedances in a coaxial tube and supplied with one and the same voltage. These connections are thus suitably arranged in such a manner, that the individual receiving electrodes are united into one single receiving cylinder, whereby the individual connections, previously required for each receiving electrode through the vacuum-tube envelope can be reduced to one single lead-in wire.

ln the abovementioned tube the `electron beam can indifferent ways be brought to make a stepping movement from receiving electrode to receiving electrode, that is, from box to box. This can, for instance, be done insuch a way, that pulses are fed in between the cathode and the electrodes, which are directly interconnected, whereby each pulse steps up the electron beam to the nearest following box. The precise course of this stepping will not be described in detail here, as it has been carefully discussed in connection with the trochotron. `It is evident, however, that tubes of the described type can be used for the counting of pulses in a pulse train.` Broadly speaking, the same rules apply to the building up of the necessary circuits for both types of tubes described, namely the trochotron and the coaxial tube. It will not be necessary to go into any further characteristics of these tubes here. Reference can be made, however, in this connection fto the magnetrons, having negative internal resistance, wherein an increase of the anode voltage within certain limits will cause a reduction of the anode current; Such a negative characteristic is also found with the trochotron and the coaxial tube, and has been discussed by H. Alfvn in a paper On trochoidal electronic beams and their use in electronic tubes (trochotrons), published in Transactions of the Royal institute of Technology, No. 22, Stockholm, Sweden, year 1948 (pages 13-16). Pulse stepping and trochotron circuits for such stepping `are described in a paper published at the same time Design of trochotron circuits by L. Lindberg (pages 39-53). In the latter paper is also described a pulse counter, including a trochotron, for the counting of Geiger-Mller pulses (pages 58-60).

The pulse circuit, shown `in Fig. l on page 39 in the above-mentioned paper by Lindberg, with. negative pulses fed in to the interconnected receiving electrodes (the plates) and with high resistances in the control electrode- (the spade) circuits, requires very well-delined pulses, the duration of which must be exactly equal to the time for switching over the beam to next box. This circuit is considered unsatisfactory. The reasons for this have not been further investigated in the above-mentioned paper; but have proved to be due to the internal tube capacitances between the different electrodes.

The present invention relates to a switching device for utilizing the described trochotrons and coaxial tubes for counting electrical pulses, whereby the harmful effects of the internal tube capacitances have been eliminated.

A switching device according to the invention for pulse counting, using vacuum tubes of the type containing an electron emitting cathode and one receiving electrode or several interconnected receiving electrodes and with` a number of control electrodes influencing the path of the electron beam within the tube, said parts being located iri a magnetic iield in which the beam moves in a plane perpendicular to the magnetic tield and in which pulses are fed between the cathode and the receiving electrode (the interconnected receiving electrodes) is principally char--L acterized by each control electrode being directly corinected via a condenser of considerably higher capacitance than the internal tube capacitances to cathode potential (ground) and/or via a condenser of the same capacity as the tube capacitances being connected to a neutraliz- V'ing voltage of the same time ratio as for the pulse sequence but counterphased in relation to this pulse sequence.

In the following the invention will be explained in connection with the attached drawings, wherein Fig. l shows a coaxial tube, connected for pulse counting, having a single cylindrical receiving electrode.

Fig. 2 shows a trochotron, connected in an analogous manner.

Fig. 3 Vshows a circuit according to the invention for eliminating the harmful internal tube capacitances.

Fig. 4 shows another embodiment of a controlling device according to the invention.

Fig. 5 shows a more detailed wiring diagram of a device for pulse counting, comprising a blocking oscillator and a' coaxial tube with Va receiving electrode, and

v Fig. 6 is a view of a tube in perspective with the eld magnet shown in vertical section.

The coaxial tube shown in Fig. l contains a cathode 21, of cylindrical form, centrally located within the tube envelope 22. The tube is permeated by a magnetic field,

Vthe direction of which (23) is perpendicular to the plane of the drawing figure. This field is conveniently derived 'from a cylindrical permanent magnet 15 such as shown in Fig. 6 coaxially surrounding the tube. The cathode is surrounded by a number of anodes or control electrodes,

designated 1-5, which via resistances l1-l5 are connected to the positive terminal of a voltage source 24, the negative terminal of which is connected with the cathode 21. The cathode and the. control electrodes are surrounded by a cylindrical receiving electrode 11, which is connected to a higher positive voltage than the control electrodes (the battery 25). A pulse-train is fed in from the

terminals

26, 27 of the transformer 28 between the cathode 21 and the receiving electrode 11. The outgoing electron beam from the cathode (which beam is not shown in the figure) will be moved forward one step for each pulse, for instance from the box formed between the control electrodes 1 and 2 to the box formed between the

control electrodes

2 and 3.

An analogous circuit usinga trochotron is shown in Fig-2. 21 and 30 in this figure represent an electron source, generating a trochoidal beam (not shown), which lby means of the control electrodes (the spades) 1-6 can be guided Vto the different boxes formed between the control electrodes, each one of which boxes contains a receiving electrode (plate) 11-16. 7 indicates the accelerating anode belonging to the electron source and 29 indicates another electrode, the so-called rail, of approximately cathode potential. The directly interconnected receiving electrodes 11-16 receive positive potential from the battery 25, the negative terminal of which is connected to the cathode 21 within the electron source. The different control electrodes are through individual resistances 41- 46 connected to a positive voltage from said battery 25. The accelerating anode is over the resistance 47 connected to a lower positive potential than the control electrodes. As in Fig. l, a negative pulse-train is via a transformer 28 fed in between the receiving electrodes Irl-16 and ground (the cathode 21).

Certain of the internal tube capacitances in Figs. l and 2 are indicated by lines of short dashes. The internal. stray capacitances are of the following kind: capacitances from control electrode to receiving electrode (35 in Fig. 1, 39, 40 and 37, 38 respectively in Fig. 2), capacitances (36) from control electrode to ground (cathode), capacitances from control electrode to control electrode, partly between adjacent control electrodes (33, 34) and partly between non-adjacent control electrodes (31, 32). Only the capacitances from the control electrode 1 have, for the sake of clarity been drawn in Fig. l, but it is evident that analogous conditions apply to all control electrodes.

The case is the same in regard to the trochotron, shown in Fig. 2.

Av brief description of the circuits shown in Figs. l and 2 follows: The negative pulse generated by the secondary winding of the transformer 23 is guided to the receiving electrode 11 and the interconnected receiving electrodes 11-16 respectively (Fig. 2), whereby the voltage between the cathode 21 and the receiving electrode 11 (and the receiving electrodes 11-16 respectively) is reduced correspondingly. If the beam is considered locked in the box, formed between the control electrodes 1 and 2, before the pulse is fed in and, therefore, directed to that part of the receiving electrode 11 which is placed in this box, the said reduction in voltage will move the beam one step to the box formed between the

control electrodes

2 and 3. On account of the reciprocal internal capacitances 35 and 37-40 respectively between the receiving electrode 11 and the control electrodes 1-5 on one hand, and the interconnected receiving electrodes 1116 and the control electrodes 1-6 on the other, the aforesaid pulse will in reality also be more or less imposed upon the different control electrodes, so that the potentials of these will be reduced. If further the potential of a control electrode is reduced, the preceding control electrode, owing to the capacitance 33 in relation to the preceding electrode, will tend to drop in potential, which may cause the electron beam previously mentioned not only to step over to the box, formed between the

control electrodes

2 and 3, but even to be moved by a single pulse to the box between the

electrodes

3 and 4, for which reason a single pulse may cause a double stepping of the beam. In unfavourable cases, even several boxes can be jumped over at the pulse stepping. When pulse counting takes place, it is necessary that the beam is properly stepped from box to box with only one step for every incoming pulse and that, in other words, the potential on one control electrode (spade) at a time must drop with sharply defined time intervals. In a tube as in Figs. 1 and 2, this time interval will be greatly influenced by the capacitances 35 and 37-40 respectively, which will therefore determine the velocity with which the potential of the spade will drop. Further, the current through the tube (the cathode current) must not be too small during the stepping movement, a condition which may easily be caused by the aforesaid capacitances between the receiving electrodes and the control electrodes. On account of these capacitances, the pulse may actually cause such a big drop in voltage at all the spades, that the potentials of these may approach zero, whereby the cathode current will be so greatly reduced, that the stepping procedure may cease.

Moreover, the internal tube capacitances for tubes of this kind are very difcult to control. Thus, they cannot be held at safe values, nor can they be kept alike owing to the dissimilarities in tube symmetry. The time for the stepping from box to box when a pulse is being received, the so called time for switch over, will not be sufciently well defined with the circuit shown in Figs. 1 and 2, partly due to the capacitances between the receiving electrode (electrodes) and the control electrodes and partly through the influence of the reciprocal control electrode capacitances. The former will greatly reduce the current at the moment for switching over, and the latter will cause risk of double stepping.

It is further evident that, with the trochotron shown in Fig. 2, the conditions at switching over will become still worse than with the coaxial tube shown in Fig. l. Owing to the unsymmetrical construction of the trochotron, the irregularities in the internal tube capacitances will be very great and, therefore, the time for switch over will vary for different control electrodes (spades). This disadvantageous action has been observed previous- 1y, and in his abovementioned paper Lindberg gives an account of how this problem has been solved with trochotorts, [wo different ,solutions to make pulse stepping possible a-re 4thereby suggested, both of which require separate plates which are not interconnected, as has been shown in Figs. 1 and 2 on the attached drawing. According to the rst solution, shown in Fig. 2 in the mentioned paper, an RC-circuit is connected to cach receiving electrode (plate) (RC-coupled, plate pulsed stepping circuit) and according to the second, the plates are divided into two push-pull fed groups (push-pull-coupled stepping circuits),

'It is, however, to be desired that the simple tube having a common receiving electrode as shown in Fig. l should be employed for these pulse counting circuits, as this tube, iirstly due to its symmetrical design, will be easier to control and secondly, because it is considerably cheaper to produce, and a great number of lead in wires (connections) through the tube bulb can be saved compared with the trochotron. The object of the present invention is to indicate the switching device, that will make this possible.

For this purpose a condenser is inserted between each control electrode and ground (cathode potential), the capacitance of which is much greater than the internal tube capacitances. A capacitive voltage division is thereby obtained, and the capacitance at each point in the tube will be essentially equal, inasmuch as it is substantially determined by the great capacitance to ground. By the introduction of such an external capacitance, the time for the switchover will naturally be changed, since this is determined by the external capacitance. The time for switch over will be longer, and the possible speed for switch over will decrease and thereby also the maximum pulse-speed, which can be used for the pulse-train, that is to be counted. The stepping (switching) time is of course determined by the product of the voltage drop, required for locking the beam in the next box, and the effective capacitance from respective control electrode to ground, and is inversely proportional to the cathode cur- -rent during the switch over.

It is evident that the insertion of a condenser between spade and ground will not only invol e advantages. The reduction of the permissible time for switch over can, however, be nullied by neutraiizing the capacitance between the spades and the receiving electrode and thereby eliminating the harmful influence of the same. This is effected by the control electrodes via a neutralizing condenser being supplied with a voltage, which is proportional to the pulse, but of reversed phase in relation to this. The harmful inliuence from the other tube capacitances will, however, not be eliminate-d, but by inserting a smaller condenser between the control elec trodes and ground, this inuence will be satisfactorily neutralized; this will result in a decrease of the time for switch over, thereby increasing the pulse speed. lf a certain reduction of the accuracy of the coupling is allowed, the condenser between the control electrodes and ground can be entirely eliminated, providing the harmful capacitance between the control electrodes and the receiving electrode is neutralized in the manner described above.

The two abovementioned couplings, having condensers in the control electrode circuits, can consequently he used both together, which seems to be most favourable, and

. also individually, for magnetic held-tubes of mentioned type, intended for pulse counting.

A coupling (switching) device is shown in 3, having condensers connecting the control-electrodes to ground. 1-5 indicate, as previously, the control electrodes, ll-45 the resistances connecting the electrodes to positive potential (battery 24), whilst 2l, ZZ, 1l indicate cathode, tube-bulb and receiving electrode respectively. The control electrodes 1-5 are through individual condensers 51-55, parallelly arranged to respective rc- `sistances i1-4SJ separately connected to cathode potential (ground), that is the negative terminal of the battery .24. The receiving electrode 11 receives a higher positive potential from the battery 2'5 through a circuit, into which the negative pulses from the transformer 28 are fed. Some of the internal tube capacitances are also designated, whereby 33, 3d, 31, 32 apply to capacitances from control electrode to `control electrode, 35 applies to capacitances to the receiving electrode and 36 to the capacitance to ground. The condensers 51-55 are much greater than the existing internal tube capacitances, and the coupling (circuit) shown in the figure will act as `a capacitive voltage divider.

lt is assumed that, when a certain pulse is received between the receiving electrode 11 and the cathode 21 the electron beam is locked to the control electrode 1, and that its current is thereby distributed between, on the one hand, this control electrode, the potential of which, by the drop in voltage in the connected resistance di, is lower than the potential on the remaining control electrodes 2-5 and, on the other hand, the receiving electrode lll, which will be hit `by the beam in the box, formed between the electrodes ll and 21. The incoming negative pulse will reduce the potential of the receiving electrode 1l, so that this becomes less positive and approaches the potential of the cathode. Even before the pulse has arrived, the range of the current-voltage-characteristic of the control electrode, within which the current to the control electrode increases with decreasing voltage, has been surpassed, and the electrode 1 carries for that reason already the maximum possible current, so that it can not receive any further portion of the cathode current. In reality, when the beam islocked to a certain control electrode, the greater part of the beam-current is picked up by the receiving electrode 11. When the potential of the receiving electrode 11 is reduced, the electron beam recedes from this electrode. Because of this, it can not nd its way to the control eiectrode 1 which, when a pulse enters between the cathode 21 and the receiving electrode 11, has greater positive potential than the latter, as the control electrode 1 has already absorbed as much of the cathode current, as it possibly can. Therefore, when a pulse arrives, the beam can not nd any other electrode than the control electrode 2. This electrode will therefore be hit by the electron-beam, which will be locked at the electrode 2, owing to its reduced potential, caused by the current passage through the resistance 42. The greater part of the electron beam will thereby be received in the box formed by the

electrodes

2 and 3, since the receiving electrode 1l will increase its potential and again take up current from the cathode when the pulse ceases. Evidently, it is important that the duration of the pulse is such, that this can take place at a time, when the capacitance S2, connected to the electrode 2 in co-operation with the resistance d2, has reduced the potential of the electrode 2 to the locking value. At this moment the electrode l ceases to carry any current and receives the same potential as the electrodes 3-5.

When the next pulse is received over the secondary winding of the transformer 28, the described process will be repeated. The potential of the receiving electrode l1 drops and the beam is forced over to the electrode 3 and the box existing between the

electrodes

3 and 4. This stepping movement of the: electron beam from box to box within the tube, will continue until the pulse-train ceases. Knowing the start and final positions of the beam, it is evidently possible to determine the number of pulses, that have occurred. Further it is evident, as pointed out before, that for the mentioned purpose it is important, that the drop in voltage on one of the `control electrodes l-S does not cause a decrease in voltage on the remaining control electrodes, which would involve the risk of double stepping. This is accomplished by the individual condensers 51-55, which are connected to ground, as `shown in Fig. 3, which condensers form a capacitive voltage divider together with 7 the internal tube capacitances. VOn account of the considerably greater capacitances of these condensers 51-55, than the internal tube capacitances, the influence of the latter can be disregarded when deciding the time for switch over from box to box. When the incoming pulse reduces the potential on the receiving electrode 11 to approximately cathode potential, the electron beam will be pushed away from this to the nearest positive electrode, that is, the control electrode 2, as described before. For a short moment, this will receive the whole part of the beam, which has been repelled from the receiving electrode 11. The cathode current will thereby charge the condenser 52 which is completely charged Aafter a certain time, which time duration is determined by its capacitance (C) and the necessary drop in tension (V) on the electrode 2 for the locking of the beam to the same, and also the cathode current (l) (the time 'TCV/I). During this time, the major part of the cathode current will be consumed for charging the condenser, and a minor part of the cathode current will `iiow through its parallelled resistance 42. When the potential on the control electrode 2 has dropped near to the cathode potential, this control electrode, having previously lreceived the whole cathode current, can no longer retain more than a small part of this current, which amount is determined by the resistance 42 and the tube characteristics. The remaining part of the cathode current must therefore try to ind a way over to another positive electrode. The control electrode 1 cannot be reached, inasmuch as a potential barrier already has been formed between it and the electrode 2. The beam must therevfore pass over to the opposite direction. The duration of the pulse is therefore so accommodated, that the mentioned pulse ceases when the condenser is completely charged, so that the receiving electrode 11 becomes more positive again than the control electrodes. Yet, in this case the pulse-length is not so critical, because the de- .vice works satisfactorily even if the pulse length should vary in the relation 1:2. The part of the cathode current, which exceeds the saturation current for the control electrode 2, will try to find a way to the receiving electrode 11 in the box formed between the

control electrodes

2 and 3. It follows from what has been said above, that the condenser 52 will be the determinating factor in the choice of pulse-length and pulse-speed.

Each control electrode in the circuit shown in Fig. 4 is connected via a condenser of about the same capacity as the internal tube capacitances to a neutralizing voltage from the pulse, having the same time-ratio as the pulse succession, but counter-phased in relation to this.

'Ihe cathode, the receiving electrode and the tubeenvelope are in this ligure, as in the preceding figures, designated 21, 11 and 22 respectively. 1-5 indicates the control electrodes which are connected by the resistances 41-45 to a certain potential. 2.4 and 25 are the voltage sources for control-electrode and receivingelectrode voltage respectively. 2.6, Z7 and 28 indicate the transformer from which the pulses arefed.

Each control electrode 1-5 is connected by an individual condenser 61-65 to a common lead, which is connected to one of the secondary terminals of the transformer.

A tap 71 on the secondary winding is connected to the common lead for the control electrode resistances 41-45 through battery 25.

Some of the internal tube capacitances are also indicated in the figure, namely the capacitances between control electrodes and receiving electrode S51-355.

The purpose of the indicated neutralizing condensers 61-65 is, as earlier mentioned, to nullify the influence .of the capacitances shown between the receiving electrode and the control electrodes. These capacitances cause, as also has been intimated in the foregoing, a decrease in the cathode current when the pulse is being fed in.

A brief description of the action of the device follows: A positive pulse-train is fed in to the

primary terminals

26, 27 of the transformer 28, whereby each pulse gives, as before, a negative pulse to the secondary winding of the transformer, which pulse will be delivered between the cathode 21 and the receiving electrode 11, thereby causing the reduced voltage between them, which is necessary for the beam to switch over.

A portion of this negative pulse will be carried over the internal tube capacitances S51-355 from the receiving electrode 11 to the different control electrodes 1-5. The potential of these would obviously be reduced on account of the pulse, by which the cathode current would be reduced, as the potentials of all anodes would thereby approach cathode potential. To counteract such an effect, the tap 71 is arranged on the secondary side of the transformer. From this tap, a pulse of the same phase and form as the positive pulse will namely be supplied to the control electrodes 1-5 over their neutralizing condensers 61-65 with the obvious result, that the potential drop, which is caused by the negative pulse, transferred from the receiving electrode 11 to the different control electrodes 1-5 via the stray capacitances 351-355, will be compensated and neutralized.

The elfect of the other harmful tube capacitances will, however, not be nullied by the inuence of the neutralizing condensers 61-65. The condensers 51-55, shown in Fig. 3, will when necessary, be used for this purpose; these condensers will reduce the inuence of all kinds of internal tube capacitances (not only the ones between receiving electrode and control electrodes), with the only inconvenience that the maximum allowed time for switch over from box to box within the tube will be reduced. In case the effect of the capacitances S51-355, indicated in Fig. 4, should be nullified by the neutralizing condensers 61-65, it is evident that in order to cancel the effect of the other internal tube capacitances, condensers 51-55 of smaller capacity can be used, whereby a higher speed for switch-over can be allowed when using the combination shown in Figs. 3 and 4, than by using only the circuit shown in Fig. 3.

A decadic pulse-counter, arranged according to this invention, is shown in Fig. 5, whereby the pulses to be Vcounted are fed in to the coaxial-tube over a blocking oscillator.

l-ltl in the igure indicate control electrodes, 11 a receiving electrode, 23 the direction of the active magnetic field, 24 a voltage source, 26 and 27 the terminals to which the incoming pulse-train is being fed in, 28 a blocking transformer, 41-50 resistances connecting respective control electrodes to ground, 5160 condensers parallelled to these resistances, 72 a buffer condenser for the

battery

24, 73 and 74 output terminals for the pulsetrain, 75 a grid-bias battery, 76 and 77 vacuum-tubes forming part of the blocking oscillator, 78, 87, 88 condensers and 79, Si), 82-86 resistances forming part of the blocking oscillator and 81 the grid of the tube 76 inthe mentioned oscillator.

The circuit shown operates in the following manner: A negative pulse-train is fed in over the

terminals

26, 27 of the blocking oscillator, the input of which is matched in such a way by the condenser 78 and the resistances 79 and 8i), that it is insensible to negative pulses below a certain amplitude (for instance 5 volts) and also to positive pulses. Only negative pulses exceeding this amplitude will consequently be fed in to the grid 81 of the iirst tube 76, working as an amplifier; this tube will amplify the incoming pulse and feed it in as a positive pulse to the blocking oscillator tube 77. This tube is normally blocked with about 20 volts grid-bias from the battery 75. The blocking oscillator is arranged to give a pulse with a length of about 0.7 micro-seconds, which over the third winding in the blocking-transformer 28, is applied to the coaxial-tube between the receiving electrode 11 and the cathode 21. The time constants in the drivingand blocking-circuits are furthermore selected in such a way, that an Vincoming pulse can only give one Vsingle outgoing pulse from the blockingoscillator, ir-

respective of the form and dimensions of the incoming pulse.

` The negative pulse applied to the receiving electrode 1l, compels the whole beam current (about 10 ma.) in the coaxial-tube to move to the next control electrode (for instance 2). The control electrode 2 drops in tension at a speed, which is determined by the beam current and the capacity of the condenser 52. When the control electrode 2 has dropped suiciently in voltage to cause stepping of the beam and its locking in next box, the pulse fed in between the receiving electrode and the cathode must cease, so that the receiving electrode again becomes positive when the beam current can divide itself between the control electrode and the receiving electrode. The effective length of the pulse is for instance at the value of 40 pf. for the condensers SL64? and a beam-current of l ma. about 0.4 micro-seconds. The device (the decade) according to Fig. 5 will deliver an output pulse from the terminals 73 and 74 for every tenth incoming pulse. This output pulse will be taken out from a capacitive tension divider, containing the condensers 55 and 555 belonging to the control electrode 5, which control electrode could be called a zero-electrode. The outgoing decade-pulse has an amplitude of about 30 volts and is negative. When the electron beam enters the box between the electrodes 5 and 6, a negative pulse will primarily be obtained at low stepping and with a resistance load, and a positive pulse of somewhat smaller amplitude when the beam leaves the mentioned box. Moreover, a smaller negative pulse (leakagepulse) will come from the output, one step before the beam enters the mentioned box, and also a small pulse, capacitively transmitted for every incoming pulse. These negative, disturbing pulses are, however, always smaller than a fourth of the main pulse amplitude. Moreover, in order to enable a decade of the kind shown in Fig. 5 to drive another decade, the input has been made insensible to small negative pulses and also to all positive pulses, as mentioned earlier.

An arbitrary number of the switching devices (decades), shown in iFig. 5, can be connected in series without any external wire connections whatever. Each decade is thereby provided with a plug at the input side and a jack at the output side, whereby the contacts in the plug correspond to the contacts in the jack. All other contacts in plug or jack within the same decade are connected mutually and with working voltages necessary for the function of the decade. In this way the last decade only, needs to be connected with a plug, receiving voltages from a power source, and the output pulse is fed in to a mechanical counter.

When using a number of interconnected decades, the zero positioning of all these decades can be performed by a common push button on the common power source, connected to the last decade. By pushing the button, the D.C. voltages to all the coaxial-tubes will be switched olf and by letting the button up, the voltage to the receiving electrodes and all control electrodes 1 4 and 6- in all coaxial tubes will increase more rapidly, than the voltage to the control electrode S and, therefore, the electron beam within all decade-tubes will be locked at the zero-electrode 5.

Indication of the position of the beam within the coaxial-tube is performed by the beam, illuminating a fluorescent screen, forming part of the tube itself. A tube designed in this manner has been suggested earlier.

The adjustable resistance in the cathode lead of the coaxial tube, designated 82 in Fig. 5, is intended for adjusting dissimilarities between different tubes. However, this adjustment is not critical, but for best function when changing tubes the resistance should be adjusted in its middle position. Such a change of tube does not generally require any adjustment.` tlf the adjusting `resistor is correctly positioned, the decade will for instance function at a supply voltage of 20G-300 volts. The voltage for the blocking oscillator must be kept between ZOO-300 volts, while the voltage tothe control electrodes and the receiving electrode of the coaxial tube can be selected between 130 and 250 volts after repositioning, of course, of the adjusting resistor. The decade-counter should function at i20% variation of the coaxial-tube voltage. The current consumption s about 10 ma. for each of the blocking oscillators and the coaxial-tube, and the heater supply is 6.3 volts and 0.75 a.. per decade.

For the device shown as an example in Fig. 5, vthe comprising components have the following values: The tube 76-77 is type 6J 6. The resistance values are: 79: 470 kiloohms, 80: 4.7 kiloohms, 83: 18 kiloohms, 84: 4.7 kiloohms, 4.7 kiloohms, 86: 4.7 kiloohms, 82: 15 kiloohms, 41-50: 180 kiloohms. The voltage of the battery 24 is 250 volts and the voltage of the battery 75 is 20 volts. The condenser values are: 78: 100 pf., 87: 100 pf., 88: 100 pf., 55: 75 pf., 555: 50 pf., 51-54 and 56-60: 30 pf. The transformer 28 has 100 turns on all windings.

It is evident, that the use of a coaxial tube, having a common receiving electrode 11 with pulsing on the latter, makes a very simple tube construction possible and also a simple and reliable driving device forl the tube. The coupling (circuit), indicated in the aforementioned paper by Lindberg, in which no condensers have been connected to the control electrodes, has on the contrary proved to be unreliable for pulse counting and caused uneven speed for switch-over from box to box on account of varying internal tube capacitances. In the circuit by Lindberg, a strong pulse is also obtained, capacitively transferred from receiving electrodes to control electrodes, which greatly impairs the function. These inconveniences will disappear entirely when loading the control electrodes with condensers according to the invention.

The invention is of course not restricted to the indicated examples of design and application, but it can be modified in several ways without exceeding the scope of the invention.

The device according to the invention is particularly adapted for counting of Geiger-Mller pulses or, in general, radio-active pulse-courses. The device is also useful for counting of for instance manufactured articles in such cases, when mechanical counters: no longer can be used because of their limited counting speed. A stopwatch formed of series connected decades according to the invention is obtained by holding the switch-over speed of the electron beam constant and continuous, for instance by control from a pulse-train produced by a crystal oscillator. The course (pulse-train), the time of which is to be determined by the electronic stop-watch, will thereby control the start and stop system of the watch. It is obvious that intervals of very short dura tion can be measured in this way.

We claim:

1. A switching arrangement capable of pulse counting comprising in combination; a vacuum tube including an electron emitting cathode, a plurality of independent control electrodes for influencing the path of the electron beam within the tube; a receiving electrode having an area thereof exposed to the cathode between each two control electrodes, the said areas being electrically connected, and means providing a magnetic field in said tube oriented to cause a beam of electrons to move in a plane perpendicular to said tie-1d in a path between said cathode and other electrodes; means to feed pulses between the cathode and the connected areas; means energizing the said connected areas at a potential positive in relation to the cathode; means energizing the individual control electrodes to a positive potential and in cluding an impedance for each control electrode; and a separate capacitance directly connected to each control .electrode and to a potential definitely related tothe cathode potential.

2. The device 'defined in claim 1 in which the last mentioned potential is cathode potential Vand each capacitance is of substantially higher value than the internal tube capacitances associated with its connected control electrode.

3. The device defined in claim 1 in which the last mentioned potential is a neutralizing one of the same time ratio as the pulse succession and counterphased in relation to the same, and each capacitance is of substantially the same value as the internal tube capacitances associated with its control electrode.

References Cited in the le of this patent UNTED STATES PATENTS Hull July 16, Langner Aug. 6, Overbeek July 30,` Desch et al. Dec. 16, Alfven et al. Aug. 14, Backmark Apr. 18, Skellett Dec. 2, Wallmark May 12, Sternbeck Nov. 17,