US2997626A - Triggered thyratron pulse distributor with standby tube - Google Patents

Description

Aug. 22, 1961 H. H. ADELAAR ET AL 2,997,625

TRIGGERED THYRATRON PULSE DISTRIBUTOR WITH STANDBY TUBE Filed Sept. 8, 195a I "4 6 F|G.2

J -P/ FIG.3

I p4 z :1 I r Y t FIG lnuenlor H.H. ADE AR J.L. MASE y ttorney United States Patent U 2,997,626 TRIGGERED THYRATRON PULSE DISTRIBUTOR WITH STANDBY TUBE Hans H. Adelaar and Jean L. Masure, Antwerp, Belgium,

assignors to International Standard Electric Corporation, New York, N.Y., a corporation of Delaware Filed Sept. 8, 1958, Ser. No. 759,564 Claims priority, application Netherlands Sept. 17, 1957 4 Claims. (Cl. 315-4545) The invention relates to a triggered pulse distributor, and more particularly to a pulse distributor for the production of phase shifted trains of pulses of equal frequency, each stage of said distributor being triggered to deliver an output pulse upon the joint occurrence of a trigger pulse, an enabling pulse and a delayed output pulse from the stage of the distributor which was previously triggered.

If such a pulse distributor operating on a ring counter basis must be designed to deliver phase shifted trains of pulses of appreciable power, for instance when it is to be used to produce pulses for the control of telephone exchanges, it may be attractive to use power gas tubes as the active elements for the distributor stages. However, such power gas tubes if they are to be exploited at their maximum power rating, must be capable of handling pulses of large amplitudes, e.g. 600 volts, and such voltage levels are generally not suited to the purpose in view, since the power should generally be delivered at a much lower voltage level, e.g. volts or so.

The object of the invention is to design pulse distributors using gas tubes and adapted to handle large amounts of power.

In accordance with a characteristic of the invention, a triggered pulse distributor of the type defined above is characterised in that each stage includes at least one power gas tube, e.g. a thyratron, working substantially at its maximum rated voltage, that said enabling pulses are applied to the anodes of said gas tubes to enable them to be fired upon the joint occurrence of said trigger pulse, derived from said enabling pulse, and of said delayed output pulse so that said trigger pulse reaches and fires the auxiliary gap of a particular gas tube, and that stepdown output transformers are connected to the cathodes of said gas tubes to reduce the pulse voltages obtained at the cathodes when the corresponding gas tubes are fired.

In this manner, the primary winding of the output transformer need not follow the steep output waveforms of the pulses applied to the anode of the power gas tube. This means that the power output transformer which has necessarily an appreciable capacitance with respect to ground, will not constitute an undesirable load for the anode pulse source, since it will be decoupled from the latter by way of the power gas tube.

In accordance with another characteristic of the invention, a triggered pulse distributor as previously defined, is characterised in that the cathode of said gas tube is connected to a first fixed DC. potential through a first rectifier and that the end of the primary winding of said transformer which is not directly connected to said cathode is connected to a second fixed DC potential through a D.C'. bias source calculated to minimize the magnetizing current of said transformer while it allows said first rectifier to clamp said cathode to said first DC. potential in the intervals between cathode pulses without creating a DC voltage component across said transformer and that said cathode is coupled to the grid circuit of said gas tube through a first D.C. restorer, whereby a floating but constant negative DC. bias is provided between said grid and said cathode and the grid limiting resistance may be of low value.

FIG. 1 is a composite schematic wiring and block diagram of a multi-stage pulse distributor according to the invention;

FIGS. 2, 3, 4 and 5 are respective wave diagrams used in explaining the operation of the system of FIG. 1.

It will therefore be appreciated that the use of a cathode output transfonmer for the purpose envisaged is even more advantageous when a DC. bias source is used to polarise the primary winding of the output transformer so that the cathode end thereof may be clamped to a fixed DC. potential without creating a D.C. voltage component across the transformer. The main advantage of this clamping arrangement is that the cathode of the gas tube is not allowed to go negative after each conduction interval, which would endanger the required de-ionization of the gas tube. With such an output cathode transformer, the required DC. bias may have one of its ends connected to a fixed D.C. supply source whereby it can be used in common for all the output transformers in the various stages of the distributor. This will be particularly advantageous when the DC. bias source is an automatic bias arrangement since the by-pass condenser may then be made very small.

Further, the floating DC. bias between the cathode and the grid of the gas tube will avoid a substantial potential difference between the cathode and grid when the potential of the former is raised upon an output pulse being produced. Otherwise, one would have to provide a large grid limiting resistor to limit grid current and this would be undesirable as it would increase the deionization time of the gas tube. Moreover, it would also delay the ionization time of the auxiliary gap.

The above-and other objects and characteristics of the invention will be better understood from the following description of a detailed embodiment thereof, to be read in conjunction with the accompanying drawing which represents one stage of a triggered pulse distributor in accordance with the invention.

Referring to FIG. 1 there are shown a number of stages designated stage No. 1, stage No. 2, and the final stage, stage n, one particular stage ST (stage No. 2) of a pulse distributor comprising n such stages connected in well known ring counter fashion. Each stage comprises a thyratron TH of which the anode is connected to terminal P through a key K. At terminal P a square waveform is shown, which may have for example a frequency of 2 kc./s., and the voltage of which may swingbetween -l30 volts and 390 volts. The cathode of thyratron TH, to which the second grid is connected, is biassed to 130 volts through the primary winding of the cathode output transformer T in series with the automatic bias network R /C which, as shown by the multipling arrow is used in common for all the n stages of the distributor. The control grid of thyratron TH is biassed to 200 volts through the grid limiting resistor R in series with resistors R and R Resistor R is a load resistor for the secondary winding of the input transformer T across the primary winding of which will be applied trigger pulses to cause the ionization of the thyratron. One secondary winding of the cathode output transformer T has been shown and one of its ends is directly connected to volts while the other end is connected to output terminal P through a rectifier W It is also connected to terminal P which leads to the input terminal of the next stage. The input terminal for stage ST which is shown, and coming from the preceding stage, is P and when a suitable positive pulse arises thereat, it will be applied through resistor R in series with a saturable reactor SR. The end of SR on the other side of resistor R is shown to be grounded through condenser C The arrangement of R SR and C constitutes a term-resonant circuit which may be suitably designed to produce a pulse across condenser C which is a replica of the input pulse to terminal P but which is delayed by a certain time which will be assumed to be equal for example to 125 microseconds. More details concerning this arrangement may be obtained from our copending US. patent application Serial No. 708,253, filed January 10, 1958, and entitled Pulse Delay Circuit.

As shown on the drawing, negative trigger pulses are regularly applied at terminal R; which is capacitively coupled through condenser C to the lower end of the primary winding of the input transformer T of which the upper end is connected to the floating end of condenser C through rectifier W The lower end of the primary winding of transformer T is biassed to 90 volts through resistor R in series with resistor R the former being shunted by rectifier W and the latter by smoothing condenser C It will be recognized that condenser C rectifier W and resistor R constitute a DC. restorer for coupling the negative trigger pulses at terminal P to the primary winding of transformer T On the other hand, resistor R shunted by condenser C constitutes an automatic bias network which rectifies the pulses appearing at the various output terminals such as P all of which are connected to this automatic bias network through an individual rectifier W so that during the operation of the pulse distributor, a potential difference of some 90 volts will appear across resistor R raising the bias of the DC. restorer to about ground potential. Such an automatic bias network is only provided for the first stage of the distributor which will be the first to be triggered when starting the operations. For the remaining stages of the distributor, the cathodes of the rectifiers such as W are directly biassed to ground potential. This automatic bias network R /C,- will permit to bring the anode of rectifier W to a potential which is under the bias potential of --90 volts by an amount corresponding to the amplitude of the negative trigger pulse e.g. 100 volts, when the latter arises. The instantaneous negative potential thus produced at the lower end of the primary winding of transformer T will be sufiiciently lower than the potential of -90 volts initially prevailing at the floating end of condenser C to initially render rectifier W conductive though no output pulse from the previous stage can yet be applied at terminal P More details concerning these particular starting circuits are shown in U.S. Patent No. 2,928,939, issued March 15, 1960.

As a result of rectifier W being made conductive, a current pulse will flow through the primary winding of transformer T and will induce a voltage, e.g. 100 volts,

across the secondary winding of this transformer which will have sufiicient amplitude to overcome the negative bias which the control grid of thyratron TH possesses with respect to its cathode, i.e. -200-(130)=70 volts. Accordingly, thyratron TH will fire and will produce a positive output pulse at its cathode which will be reproduced across the secondary winding biassed to ---90 volts to reach the output terminals P and P It will be noted that the cathode of thyratron TH is coupled to the lower end of resistor R through condenser C and that resistor R is shunted by rectifier W These last three elements also constitute a DC. restorer which permits to keep a constant floating bias between the cathode of the thyratron and its grid circuit. Otherwise, if the control grid were not positively tied to the cathode in this manner, resistor R would have a much larger value in view of the high voltage pulses appearing at the cathode of the thyratron. This would substantially hinder the de-ionization as well as the ionization of the thyratron. With a mere capacitive coupling between the cathode and the grid of the thyratron, the voltage capacitively coupled into the grid circuit would be subjected to exponential variations unless the time constant is very large. Moreover, the grid bias of 200 volts would have to be raised to take into account the fact that the potential at the left-hand side of condenser C would go below the fixed bias level by an amount equal to aE where a is the time duty ratio of the pulses at the cathode and E their amplitude. Variations of aE would therefore also cause variations in the floating DC. bias between cathode and grid. When the rectifier W is added to prevent negative voltage excursions below 200 volts, between successive pulses, the cathode will be at l30 volts while the grid will be biassed to -200 volts to maintain the desired bias of -70 volts. During the pulses, the full amplitude of the voltage step of +520 volts (abstraction being made of the small plate-to-cathode potential) at the cathode of the thyratron will be injected into the grid circuit to raise the absolute value of the bias of the grid but maintaining its relative value with respect to the cathode potential, at 70 volts below the latter, i.e. +390 70=320' volts.

Between the pulses, the cathode of the thy-ratron is at 130 volts because it is clamped to that potential through rectifier W in series with the winding of the common relay Tr used for all the n stages of the distributor. Across resistor R a potential of aE volts is developed to avoid that a DC. voltage component should arise across the primary winding of transformer T which would tend to saturate the latter. At the same time, this automatic bias across resistor R still permits rectifier W which is blocked during a positive pulse at the cathode, to be conductive between successive pulses. Additional information concerning this particular output transformer circuit may be found in Belgian Patent No. 571,- 265, issued March 17, 1958. Relay Tr will regularly receive current during the intervals between successive pulses and may therefore be arranged to normally remain operated, If it should release, this may therefore be used to give an indication that the rectifiers such as 'W no longer conduct between pulses, which might for example arise if the voltage across R becomes larger than the value of aE. Then, the value of R could be adjusted to decrease the voltage across this resistance to the optimum value.

During the operation of the pulse distributor, the first stage shown will function in the same way as the other stages, i.e. an output pulse from the previous stage and appearing at terminal P will be necessary to produce a delayed pulse raising the anode voltage of rectifier W In such a case, although the negative trigger pulse capacitively fed from R; will be based on a level substantially equal to ground by virtue of the clamping action of rectifier W and the potential difference of about volts across R the potential difierence across rectifier W will be sufficient to unblock it and induce a positive trigger pulse into the grid circuit of the 'thyratron to ionize the latter. Deenergisation of the thyratron TH will of course occur upon the disappearance of the anode pulse at terminal P The use of the DC. restorer C R W; for injecting the trigger pulses permits to base these on a fixed D.C. level which will be found useful in connection with the gating rectifier W The input transformer T is a convenient way to isolate the two D.C. restorers.

Whenever an absolute reliability of operation is re quired, for example for use in telephone exchanges, failure of a thyratron should not cause a failure of the distributor. In the Belgian Patent No. 523,622 (C. Weill et al., December 9, 1918), it has already been proposed to associate a spare thyratron with a normally working thyratron, the latter being normally given preference in operation due to a lower bias. Since the pulses used to fire the thyratrons must necessarily have a finite slope, the thyratron with the small bias will be the 'first to ionize. As soon as it is ionized, use of 'a common load impedance will readily prevent the ionization'o'f the spare thyratron which will only fire if the normal tube fails to do so.

Such an arrangement is however quite awkward to apply to the circuit shown since one would have to go to the expense of duplicating the floating biasses arrangements. Instead, the figure shows that the spare thyratron TH which has its cathode and its second grid commoned to those of thyratron TH, while its anode is commoned to that of TH through the primary winding of transformer T has its control grid connected to the control grid of TH through a simple delay circuit comprising resistor R between the two control grids and shunt condenser C between the control grid and the cathode of TH. In this manner, the two thyratrons may be blessed by a common arrangement and thyratron TH will normally never fire since thyratron TH is given preference. Moreover, one is no longer dependent on the shape of the pulses applied to the control grid of the normal thyratron TH. Should thyratron TH fail to be ionized within the time delay provided by R /C TH will fire instead. This delay produced by R /C will of course be chosen quite small with respect to the duration of the output pulses so that if the spare thyratron comes into action, the length of the output pulses will hardly be reduced.

Transformer T will permit to induce 'a voltage between terminal P and P upon the firingo-f TH. A non-urgent alarm can therefore be given to the maintenance staff. Preferably, terminals P and P' will be connected to a relay detecting arrangement through a rectifier network (not shown) so that it is only the repeated firing of the spare thyratron TH which will be detected, and odd and temporary failures of TH to fire first remaining unnoticed. If TH fails key contact K may disconnect it from the circuit.

The waveforms shown indicate that the trigger pulses at terminal P lag by A of a period behind the positive flanks of the anode pulses at P Therefore, the ferroresonant delay device may conveniently be arranged to produce a delay of /2 a period. The delay from one stage to the next should obviously be larger than the delay of the trigger pulses, since otherwise the delayed pulse would end before the ocurrence of the trigger pulse. On the other hand, with the ferro-resonant delay circuit, the delayed input pulse cannot start after the output pulse from the previous stage has subsided. Therefore, the limits are in the present case: A of a period and of a period which limits justify the value shown on the figure.

It will be remarked that the gating arrangement for the trigger pulses includes the rectifier W whereby, since the load at terminal P is also connected through a rectifier W and since a rectifier W is also used to produce the automatic starting bias, the load seen at the cathode of the thyratrons is unidirectional, being chosen, to be quite high in the intervals between the pulses and thereby limiting the value of the magnetising current through the transformer.

While the principles of the invention have been described above in connection with specific apparatus, it is to be clearly understood that this description is made only by way of example and not as a limitation on the scope of the invention.

What We claim is:

-1. A multi-stage triggered pulse distributor for gener ating phase-shifted trains of pulses of equal frequency,

6 comprising means for connecting each stage to a source of triggering pulses, to a source of enabling pulses, and to delayed pulses from an immediately preceding stage, a power gas tube in each stage having an anode, a control electrode, and a cathode, means for applying the said enabling pulses to the said anode of any said. tube simultaneously with the joint application of triggering pulses and delayed pulses to the said control electrode thereof to ionize the last said tube to generate a voltage output pulse, each of said tubes having an output transformer and first and second clamping means associated therewith, a primary winding on each said output transformer, means for connecting each said primary winding between the said cathode of its associated tube and a first direct current potential, means for connecting each said first clamping means between the said cathode of its associated tube and the said first direct current potential, to clamp said cathode to said first potential, and means for connecting each said second clamping means between the control electrode of its associated tube and a second direct current potential to reduce the pulse voltage obtained at the cathode when the associated tube is fired and to provide a constant negative direct current bias between the said control electrode and said cathode of said associated tube.

2. A multi-stage triggered pulse distributor as claimed in claim 1, wherein each said second clamping means includes a rectifier in .circuit with the said control electrode of its associated tube and includes a first directcurrent restorer between said control electrode and said cathode of its associated tube.

3. Ina multi-stage triggered pulse distributor as claimed in claim 2, each of said tubes having an input transformer associated therewith, each said input transformer having a primary and a secondary winding thereon, the said means for applying said delayed pulses to said control electrode of any tube including a second rectifier connected to one end of the said primary winding of the associated input transformer, the said means for applying said triggering pulses to said control electrode of the last said tube including a second direct-current restorer connected to the other end of the primary winding of the last said input transformer, and means connecting the secondary winding of the last said input transformer between the said control electrode of the associated tube and the associated first clamping means to change the conductivity of the last said rectifier to generate a pulse across the secondary of the associated input transformer.

4-. In :a multi-stage triggered pulse distributor as set forth in claim 1, each of said gas tubes having a second gas tube associated therewith with each second gas tube including an anode, a control electrode and a cathode, means for connecting the said anodes and said cathodes of the second tubes to respective anodes and cathodes of their said associated tubes, delay means for each stage including a resistor and a condenser, means including the said delay means for connecting the said control electrode of each second gas tube to the control electrode of its associated gas tube, whereby the said second gas tube of any stage is permitted to ionize only upon failure of the associated tube to ionize.

Wold July 17, 1923 Posthumus May 23, 1950

Images