US2549831A

US2549831A - Conversion system

Info

Publication number: US2549831A
Application number: US28304A
Authority: US
Inventors: Richard L Longini
Original assignee: Westinghouse Electric Corp
Current assignee: CBS Corp
Priority date: 1948-05-21
Filing date: 1948-05-21
Publication date: 1951-04-24
Anticipated expiration: 1968-04-24

Description

April 4, 1951 R. LONGINI 2,549,831

CONVERSION sYs'fEM Filed May 21, 1948 4 Sheets-Sheet l WITNESSES: INVENTOR BC/mrdLAong/ru.

- BY EM HA P: M

ATTORNEY April 1951 R. L. LONGlNl 2,549,831

CONVERSION SYSTEM Filed May 2]., 1948 4 Sheets-Sheet 2 I Phase filw/fer 1 445 L y W l WWW WWW w '3 INVENTOR lab/10rd! Z 0/2977.

ATTORNEY April 1951 R. L. LONGINI CONVERSION SYSTEM 4 Sheets-Sheet 5 Filed May 21, 1948 iNVENTOR WITNESSES: myfl f a, W yr.

V Y m M m M. hm w M April 24, 1951 LONGINI CONVERSION SYSTEM 4 Sheets-Sheet 4 Filed May 21, 1948 INVENTOR /chara A orig/) I l 5w/I'C/7 6/ajea WITNESSES: 54W.

l w Jr? ATTORNEY Patented Apr. 24, 1951 CONVERSION SYSTEM Richard L. Longini, Pittsburgh, Pa., assignor to Westinghouse Electric Corporation, East Pittsburgh, Pa., a corporation of Pennsylvania Application May 21, 1948, Serial No. 28,304

My invention relates to an electronic control system, and it has particular relation to a control system useful in resistance welding.

In one of its specific aspects, my invention involves a welding system in which single phase alternating current for welding is derived from a polyphase source. In control systems constructed and operated in accordance with the prior art, of which I am aware, current of one polarity is supplied through a first plurality of valves to the welding transformer and current of the opposite polarity through a second plurality of valves. By operation of a control system these two pluralities of valves are rendered conductive alternately at the periodicity of the single phase current. During the conductive intervals of any plurality, each of its valves conducts in its turn. At the end of a half period of the single phase, one of the valves or the first plurality is conductive. The first valve of the other plurality is now rendered conductive. To avoid a short-circuit the last valve of the first plurality should be non-conductive before the first valve of the second plurality becomes conductive. The last valve of the first plurality becomes non-conductive when the potential between its anode and cathode becomes zero or negative. Since the inductance of the welding transformer is considerable, the source voltage corresponding to the last conductive valve reaches a negative magnitude while the last valve is still conductive. The selection of the valves and the timing of their conductivity in such manner as to take into account the reactive overlap of conductivity is not practicable. source potential, in the material to be welded, in the extent to which the material is inserted in the throat of the welder interfere with the preset operation of a selection and timing system and damaging short circuits result.

It is, accordingly, an object of my invention to provide a protective circuit for a polyphase to single phase valve conversion system which will prevent the power source from becoming short-circuited.

Another object of my invention is to provide a polyphase to single phase valve conversion system in which short-circuiting of the supply by the valves will not occur.

A general object of my invention is to provide a new and improved control system for changing polyphase alternating current to single-phase alternating current.

A further Object of my invention is to provide a new and improved circuit arrangement for changing polyphase alternating current of one frequency to a single-phase alternating current of a lower frequency.

A further object of my invention is to provide a new and improved circuit for changing a poly- Variations in the 18 Claims. (01. 315-138) V phase alternating current of one frequency to a single-phase alternating current of a lower frequency in which the kilovolt-amperes demand will be substantially lower than for prior systems.

More specifically, it is an object of my invention to provide a novel control system useful for resistance welding in which single phase current is supplied from a polyphase supply.

In the prior art systems, difficulty has also been experienced with the timing circuit which controls the conductivity of the first and second pluralities of valves. I found that the timing oscillations which these circuits produce vary in frequency according to the length of time during which the circuit operates. This variation produces a corresponding variation in the load current. Since the frequency of the load-current from such a timer is dependent upon the time required for the grids of each of its plurality of valves to reach a firing potential, the frequency is also dependent on the lowest potential reached by the grids. In this type of timer the grids of its plurality of valves will not begin to rise from a uniform low potential until some time has elapsed. Further, when the control switch is opened to discontinue the supply of load current, the timer stops at a random instant in its output period. The operation generally terminates with a voltage remaining in the load circuit. This operation would leave a load transformer saturated.

It is, accordingly, a further purpose of my invention to provide a timing circuit which shall produce timing oscillations of uniform frequency throughout its operation.

Another object of my invention is to provide a timing circuit which, when the timing is to be discontinued, shall supply no appreciable timing potential.

In accordance with my invention, the conductivities of the valves of each of the pluralities are, in fact, mutually controlled from auxiliary anode circuits of the valves of the other plurality. So long as a valve is conductive its auxiliary anode circuit carries current. A potential derived from this current blocks the conductivity of any valve in the other plurality.

In accordance with my invention a timer is provided in which the period of the timing current is dependent on the time required for the grids of its plurality of valves to reach a firing potential, and each grid rises through the same potential difference throughout the operation of the timer.

The novel features of my invention are set forth with more particularity in the accompanying claims. The invention itself, however, with respect to the details thereof, together with its additional objects and advantages, may be better understood from the following description of a Fig. 3 is another series of graphs illustrating the operation or" the system shown in Fig. 1.

In the apparatus shown in Fig. l, a material 4 to be welded receives energy from three-phase supply lines 5 through a three-phase supply transformer I. The supply transformer I has an output terminal 9 for each phase and a common return terminal I5. Three main electric discharge valves II, I9 and 2| of the arc-like type, such as ignitrons, are provided to correspond with the three-phase voltages, each main valve having an anode I I connected to the corresponding output phase terminal 9 of a supply transformer I and a cathode 23 connected through the primary 25 of the welding transformer 21 to the common return terminal I5. These valves II, I9 and 2| conduct current of one polarity through the welding transformer.

A second group of three main electric discharge valves 29, 3|, 33 of the arc-like type, such as ignitrons, are provided to conduct current of the opposite polarity. Each of the latter main valves has a cathode 35 connected to the corresponding output phase terminal 9 of the supply transformer I and an anode 31 connected to the common return terminal I5 of the supply transformer I through the primary winding 25 of the welding transformer 21.

The sequence and time of firing of the six ignitrons II, I9, 2|, 29, 3| and 33 is controlled by a control circuit comprising six auxiliary valves I5I, I51, I59, III, I13 and H5 of the arc-like type, such as thyratrons. These auxiliary valves are supplied through a control transformer I45 and a phase shifter M! from the main supply 5. The conductivity of the six auxiliary thyratrons is controlled from a multivibrator.

The action of the control circuit is such that first the three ignitrons II, I9, and 2| become conductive in succession, passing current of one polarity through the welding transformer for a preselected interval of time; then the three ignitrons 29, 3!, and 33 become conductive successively, passing current of the opposite polarity through the welding transformer for the same interval of time. Thus the welding transformer receives single-phase alternating current of the desired frequency.

In the multivibrator circuit the positive terminal 4| of a first source 43 of direct current shown symbolically in Fig. lb as a battery, is connected through resistors 45 to the anode 4! of each of three valves 49, 5|, 53 of the arc-like type, such as thyratrons. The cathode 55 of each thyratron is connected to the negative terminal 51 of the first source 43 of direct current which is at ground potential. Prior to initiation of operation of the multivibrator, a heating current is supplied to the filaments of the various valves and tubes of the entire welding circuit by conventional means which, for purposes of clarity in the drawings, are not illustrated.

The grid 5| of the second thyratron 53 is connected through an impedance 89 to the anode 41 of the third thyratron 49. The positive terminal 41 of the first thyratron 5| is connected through a voltage divider to the negative terminal 65 of the second source 61 of direct current,

shown in Fig. 1b as a battery. A central point 63 in the voltage divider is connected through a resistor 89 to the grid of the third thyratron. The positive terminal 41 of the first thyratron 5| is connected through a capacitor II, a diode tube I3, and a second capacitor 15 to ground. The capacitor II is connected through a resistor I9 to ground.

When the first thyratron 5| is in a non-conductive state, the positive terminal 48 of the voltage divider is at a high potential of, for example, 150 volts. This produces a positive bias on the grid of the third thyratron 49 sufficient to allow the third thyratron 49 to fire if the first thyratron 5| were not conductive.

At the beginning of operation the control switch 59 is closed causing the first thyratron 5| to fire. When the first thyratron 5| first becomes conductive, the potential of its anode 4'! is decreased to a magnitude equal to the voltage drop of the thyratron. If the supply 43 provides 150 volts, the decrease is of the order of 142 volts. The value of the first and second resistors 8| and 83, respectively, of the voltage divider and the two condensers II and I5 is such that the anode of the diode tube I3 which is prevented from changing potential rapidly by the presence of the second condenser I5 is at a higher potential than the cathode II which is connected through a first capacitor 'II to the anode 41 of the first thyratron 5|. Since there is a potential difference of the proper polarity across the diode tube I3 sufficient to cause conduction, the diode tube I3 begins to conduct current as soon as the first thyratron 5| becomes conductive. Thus, as soon as the first thyratron 5| becomes conductive, a surge of current passes through the diode tube I3 causing the anode 85 of the diode tube I3, and the grid of the third thyratron 49 connected to it, rapidly to reach an equilibrium potential determined by the capacity of the first and second capacitors II and I5, respectively.

The capacity of the first and second capacitors II and I5 and the first and second resistors BI and 83 is such that the equilibrium potential at the point 85 between the diode tube I3 and the second condenser I5 is the same as the equilibrium potential between the first and second resistors 8| and 83. For this to be the case, the following relation must hold approximately:

lee C, R E

where Cz=capacity of the second capacitor I5 C1=capacity of the first capacitor II E1=vo1tage between the anode 41 of the first thyratron 5| and ground when the thyratron 5| is conductive E2=voltage across the second battery 61 Ri=resistance through the resistor 8| connected to the anode of the first thyratron Rz=The resistance through the resistor 83 conrtiected to the negative side of the second batery These equations state the desired relationship only approximately, because they make no a1- lowance for either the resistor I9 connected between the'cathode II of the diode tube I3 and ground or the diode tube I3 across which a potential exists, but it gives approximately the relationship required. If the two resistors 8| and 83, the two condensers II and I5 and the two given above, no current will fiow between the central point 63 in the voltage divider and the point 85 between the two condensers. The grid 69 of the third thyratron 49 will then remain at one potential until the voltage E1 is changed.

When it is desired to begin the actual welding operation, the starting switch 81 is closed. This starting switch 87 connects the grid 6| of the second thyratron 53 through an impedance 89 to the anode 4! of the now non-conductive third thyratron 49, thus raisingthe potential of the grid 6| to a value sufiiciently high to cause the second thyratron 53 to fire. The second thyratron 53 then charges a capacitor 9| toextinguish the first thyratron 5 After the first thyratron 5| is extinguished, the

potential difference across the first thyratron 5| increases greatly. Its first capacitor H ischarged through a resistor 19 and its second capacitor 15 is charged through the resistors 8| and 83 of the voltage divider. A certain time delay determined bythe time constants of the charging networks is involved in this charging process. During the time that the capacitors H and 15 are being charged, the second thyratron 53 continues conductive. As soon as the capacitors H and 15 have charged suificiently to raise the potential of the grid in the third thyratron 49 to a high enough value to allow the third thyratron 49 to fire, the third thyratron 49 fires, charging capacitor 93 to extinguish the second thyratron 53.- The grid circuit of the first thyratron 5| is connected to the anode circuit of the second thyratron 53 in the same manner as the grid circuit of the third thyratron 49 is connected to the first thyratron 5|. The same voltage, capacity, resistance relations hold in the grid circuit of the first thyratron 5| as previously described for the circuit in the grid of the third thyratron 49.

When the second thyratron 53 first becomes conductive, it lowers the potential of the control grid on the first thyratron 5| in the same manner as the first thyratron 5| had previously lowered the potential on the control grid of the third thyratron 49. When the third thyratron 49 becomes conductive, causing thesecond thyratron 53 to become non-conductive, the rid potential on the first thyratron 5| begins to rise toward the critical value required to allow the first thyratron 5| to become conductive after a time interval .predetermined by the time constant networks in the circuit of thyratron 53, the first thyratron 5| is rendered conductive. The third thyratron 49 is now extinguished by the charging of the capacitor 95 connecting the anodes 41 of thyratrons 49 and 5| respectively. With the starting switch 8'! closed, the control grid 6| on the second thyraton 53 is always at the potential of the anode of the third thyratron 49. The control grid of the second thyratron 53, therefore, immediately rises to the potential of the anode of the now non-conductive third thyratron 49 and the second thyratron 53 fires immediately after the third thyratron 49 is extinguished. The second thyratron 53 extinguishes the first thyratron 5| through the capacitor 9| connecting their anodes 41.

The sequence of operation continues as described until the starting switch 81 is opened. The startin switch 81 will be manually opened at a random instant in a period of the multivibrator period. Since the opening of the starting switch 81 serves only to prevent the second thyratron 53' from becoming conductive and does not affect the operation of the other thyratrons, the sequence of operation continues until the time when, with switch 8'! closed, the second thyratron would become conductive. The second thyratron 53 cannot become conductive because switch 81 is open and the change in the conductivities of the valves is stopped when the first thyratron 5| becomes conductive. The latter continues conductive until the control switch 59 is opened.

The voltage relationships which result from this sequence of operations can conveniently be described with reference to Fig. 3(a), (b) and (c). In these graphs the output voltage of each of the three tubes, the voltage at all times across each of the tubes, and the grid potential on each tube are plotted as a function of time. It must be realized that, for the sake of clarity, no minor variations of potential are shown in these graphs. Voltage is plotted vertically and time horizontally. The time axis for each curve is the same, so that directly above a point representing a certain time on curve 0 are points on curves 0. and b which represent .the same instant of time. The voltages relating to the first thyratron 5| are shown in Fig. 3(a); those relating to the second thyratron 53 in 3(b),.and those relating to the third thyratron 49 in 3(0).

When the control switch 59 is closed,'the grid potential of the first thyratron 5| represented by curve 91 is high enough to permit firing, and the first thyratron 5| becomes conductive. The output voltage from the first thyratron 5| represented by curve 99 becomes high and the voltage across it represented by curve ||l| becomes low.

The output voltage (curves |93 and I95) from the second and third thyratrons 53 and 49 is zero. The voltage curves I0! and I99 across the second and third thyratrons 53 and 49 is high because they present an open circuit.

At the instant I I when the starting switch 81 is closed, the grid potential (curve N3) of the second thyratron 53 is raised to its critical value I |5 and the second thyratron 53 fires. The output voltage (curve I93) which is measured across resistor 45 from the second thyratron 53 rises to a high value. The voltage (curve I01) across the second thyratron 53 drops to a low value momentarily depressing the voltage across the first thyratron 5| to a negative value, and extinguishing the first thyratron 5|.

While the second. thyratron 53 is conductive, the grid potential (curve I 9) of the third thyratron 49 is rising toward its critical firing potential- When the potential (curve I IQ) of the grid 69 of the third thyratron 49 reaches its critical potential H5, the third thyratron 49 becomes conductive. The voltage (curve |2|) across the third thyratron 49 drops to a low value and momentarily renders negative the voltage (curve I91) across the second thyratron 53, extinguishing the second thyratron 53.

While the third thyratron 49 is conductive, the grid potential (curve 91) of the first thyratron 5| is rising toward its critical firing potential 15. When the potential (curve 91) of the grid of the first thyratron 49 reaches the critical value N5, the first thyratron 49 becomes conductive. The voltageacross the first thyratron 49 drops to a low value extinguishing the third thyratron 5|. When the third thyratron 5| is extinguished, the potential (curve N3) of the control grid 6| of the second thyratron 53 quickly reaches its critical potential H5, and the second thyratron 5| becomes conductive.

This sequence of operation continues untilthe instant I22 when thestarting switch 81 is opened. The opening of this switch acts only to lower the potential (curve I'I3') of the control grid SI on the second thyratron 53 sufficiently to prevent thesecond thyratron 53 from firing. A shownin Fig. 3', the sequence of operations continues until it is: again time I24 for the second thyratron 53 to fire. At that time I23, the second thyratron 53 does not fire and extinguish the first thyratron 49 as it would. have done in. the normal sequence of operation, but the'first thyratron 49 continues conductive as long as the potential I9I exists across it. Operation will always end with the first thyratron conductive.

On the basis of prior art a multivibrator would be designed using only two thyratrons, both of which would be connected to output leads. In such an arrangement operation of the multivibrator would generally be terminated while one thyratron is conductive, and a potential difference would remain across the output leads. The second thyratron in my multivibrator is not connected to the output leads, and operation ends with zero potential across the output leads.

As shown in Fig. 3, the voltage output (curve I03) of the second thyratron 53 and the voltage output (curve I95) of the third thyratron reach high values alternately, as the third thyratron 49 and the second 53 respectively conduct except for negligible intervals during which the first thyratron is conductive. The current output of the second and third thyratrons 53 and 43 provides control voltages for the two grids I25 and I21 of a control tube I23 (Fig. lb). Potential difierences are impressed between the anodes I3I and I35 and the cathodes I29 of the control tube I23, through resistors I65 and I31, respectively, from a source I33 shown symbolically as a battery. When current is flowing through the second thyratron 53 in the multivibrator circuit, the second control tube grid I21 which is connected to the anode 41 of the second thyratron 53 of the multivibrator circuit becomes negative with respect to the cathode I29 of the control tube by the drop across resistor 45. This bias prevents current from flowing between the second anode I3I and the cathode I29 of the control tube. The first grid I 25 of the control tube I 23 which is connected to the anode 41 of the-third thyratron 49 of the multivibrator circuit is at cathode potential. Current flows between the cathode I29 and the first anode I35 of the control tube I23 through resistor I31.

The negative terminal of the first resistor I31 of the control circuit is connected to the common return terminal 54! of the first secondary I43 of a three-phase transformer I45. The output line I49 of the first winding of this three-phase secondary leads to the grid I52 of the first thyratron iI of the control circuit. Phase terminals I53 and I55 of the second and third windings of this secondary I43 lead to the grids I58 and I6I of the second and third thyratrons I51 and I59, respectively. Thus the potential across each winding of this three-phase secondary I43 in part determines the grid bias of one of the first three thyratrons I52, [BI and IE3 of the control circuit. When current is flowing through the first resistor I31 of the control circuit, the potential difference across resistor I31 renders the grids I52, I53 and I5I of the thyratrons highly negative with respect to their cathodes, and the voltage in the first secondary I43 of the transformer I45 is insufllcient' to permit the three thyratrons I5I, I51 and I59 to fire.

When the third thyratron 49 of the multivibrator circuit is rendered conductive, the potential difference between the cathode I29 and the first grid I25 of the control tube I23 is decreased so that current flow is blocked between the cathode I29 and the first anode I35' of the control tube I23. The second grid I21 of the control tube I23,.now being at cathode potential of the control tube I23, permits current to flow between the cathode I29 and the second anode I3'I of the control tube. The'negative side of the second resistor I65 is connected to the neutral conductor I61 of the secondary I69 of the three-phase transformer I45. The phase terminals of the windings of the second secondary I69 are connected to the grids of the fourth, fifth and sixth thyratrons I1I, I13 and I15; respectively, in the same manner as the phase terminals of the first secondary I43 were connected to the grids I52, I58, and NH of the first threethyratrons. Analogously to the circumstances when the current flows in the first network including the first secondary I43 of the control transformer I45, current flow in'thesecond network including the second secondary prevents the fourth, fifth and sixth thyratrons I1I, I13, I15, respectively, from firing.

The anode I54 ofthe first thyratron I5I is connected through acapacitor I11, a peaking transformer I19 and one winding- I9I of the third secondary I83 of thetransformer I45 to the common-line I39'joining the cathodes of the six thyratrons. The other five thyratrons are similarly connected through their separate capacitors, peaking transformers and windings of thethird secondary I83.

Voltage of the proper polarity impressed through the output line I49 of the first winding.

of the firstsecondary I43 of the control transformer I45 will permit the first tube I52 to fire. While current is flowing in the second network causing the grids I52, I58 and I59 to be biased more positive, the voltage across the first winding I8I of the third secondary I83 of the control transformer I45 will, at some time, be of the propermagnitude relative to the cathode line I39- to provide the second and last requisite to allow the first thyratron I52 to fire. When the first-thyratron I52 fires, it sends a pulse of current charging its capacitor which then discharges through the primary IBO- of peaking transformer I19. While current is flowing in the second control network, the'fi'rst three thyratrons I5I, I51 and I59 fire in. rotation, each thyratron sending a pulse through its capacitor to its peaking transformer I;

When the first control circuit is conductive, the voltages in the second secondary I69 of the control transformer permitthe fourth I1 I, fifth I13 and sixth I15thyratrons to fire in sequence, each sending a pulse through the primary of its pulsing transformer;

The instantaneous presence of high voltage through a peaking transformer provides one condition that' the corresponding ignitron become conductive. The ot'her conditions that'must be met before the respective thyratrons become conductive will be described with reference to Fig. la.

Associated with each ignitron is a thyratron which becomes conductive at a predetermined time thus starting current flow in the corresponding ignitron. Thegrid I85 of the first thyratron I 81 is connected through a grid resistor I89, the secondary I82 of the peaking transformer I19, the secondary I90 of a firing transformer I9I, a source of direct current, depicted in the drawing as a battery I93, to the common line I95 joining the cathodes 23 of the first three ignitrons Il, I9 and 2|. The primary I92 of the first transformer I9! is connected in series with a first secondary winding I91 of a fourth three-phase secondary 225 of the control transformer I45.

In a similar manner the grids of the second and third'thyratrons I84 and I86, respectively, are connected through their grid resistors, their peaking transformers, their firing transformers, the primary of which is connected in series with a corresponding phase winding of the fourth secondary 225 of the control transformer I45, and through a common supply battery I93 to the line I95 joining the cathodes of the first three ignitrons I1, I9 and 2I.

A second transformer 20I is provided having three windings on a single core. Each of its three windings has between its terminals the potential difference that exists between the auxiliary anode and the cathode of the fourth, fifth, or sixth ignitron, 29, 3| and 33, respectively.

Thus, if the fourth, fifth r sixth ignitrons 29, 3I, 33, respectively, is conductive and current can flow between the auxiliary anode 293 and the cathode 35 of any of these ignitrons, one of the windings of the second transformer 20I will be shorted causing a short to exist in the other two windings of this transformer 2!".

The first winding 295 of the second transformer 20| is connected through dry rectifiers 29'! to each of the firing transformers I9I, 209, and 2H in the firing circuit of each of the first three ignitrons. The rectifiers 29! are connected so as to conduct electron current from the windings of transformer 29I. A short across the latter windings in effect is a short across the primary of each firing transformer I9I, 299 and ZII. The rectifiers, on the other hand, block the flow of current from the firing transformers I 9|, 299, 2II associated with any ignitron through the firing circuit of another ignitron. If any winding of the second transformer 29I is shorted, it will cause a short circuit to exist in the primary of the firing transformer associated with each one of the grid circuits of the thyratrons which control the conductivity of their respective ignitrons. The circuit parameters are so arranged that the first ignitron I! cannot become conductive unless its peaking transformer H9 is receiving a pulse voltage and its firing transformer I9I is impressing a voltage of the proper magnitude and polarity.

The circuit arrangement of the last three ignitrons 29, 3I and 33, differs from the circuit arrangement of the first three ignitrons I'I, I9 and 2| because the anode 31 of the last three ignitrons is connected to a common line I95, whereas, the cathodes 23 of the first three ignitrons were connected to this common line.

The grid 2I3 of the thyratron 2I5 which controls the conductivity of the fourth ignitron 29 is connected through the peaking transformer 2H, a biasing source of direct current, shown in the drawing as a battery 2I9, and a firing transformer secondary 222 to the cathode of the first thyratron 2I The primary 22:3 of the firing transformer 22I is connected in series with the fourth winding 223 of the fourth secondary 225 of the control transformer I45. anode 221 and the cathode 23 of each of the first three ignitrons I'I, I9 and 2I is also Connected across the primary of th firing transformer 22I.

The following are simultaneous conditions which must be fulfilled before the fourth thyratron 2I5 will fire causing the fourth ignitron 29 to become conductive: the fourth peaking transformer 2 I! and the fourth firing transformer 22I must impress voltage of the properpolarity and proper magnitude to overcome the negative bias produced on a fourth control thyratron 2 I 5 by a battery 2 I9. For the firing transformer 22I to impress a voltage of the proper magnitude, the following condition must be fulfilled. At the time chosen for the fourth ignitron 29 to fire, the sixth Winding 223 of the fourth secondary 225 of the control transformer I45 must develop a high voltage of a proper polarity. If current is flowing in one of the ignitrons I'I, I9 or 2I, its auxiliary anode will conduct electron current from its cathode and the fourth firing transformer 22I will then be virtually short-circuited. Under such circumstances the ignitron 29 will fail to conduct.

If the ignitrons II, I9 and 2I are non-conductive and conditions for firing ignitron 29 are estabtransformers associated with the last three ignitrons and acommon line 229 so that these transformers can receive energy from the control transformer I45 but can supply firing potential only in the firing thyratron circuit in which they are connected,

The current and voltage relationships and the conditions which must be met before each of the ignitrons fires can be more fully described with reference to the graphs in Figure 2(a), (b) and (c). In each of the three graphs time is plotted horizontally and voltage or current vertically. Points aligned vertically on the three graphs represent an identical instant. In Fig. 3(a) curves 239 represents the current flow through the weld: ing transformer and curve 23I represents the source potential.

When current is being conducted to the welding transformer in one direction as in the lefthand of the drawing, the voltage curve 23I is shown as being above the neutral axis, and in the second /3 of the drawing in which the welding transformer is receiving current of the opposite polarity, the voltage output of the ignitrons is shown as being below the neutral axis. Thus the output of the ignitrons is depicted as seen from the welding transformer. If the current and voltage were shown as seen from the power transformer, both the current and voltage output of the ignitron rectifiers would be graphed above the neutral axis. The present drawing takes cognizance of the circuits external to the rectifiers which cause the last three ignitrons to present a negative current to the welding transformer.

The voltage wave of the one phase 233 which is rectified by the first and fourth ignitrons I! and 29 is drawn completely as seen from the power transformer. If the other two phases had been drawn completely in this manner, the drawings would have been unduly complicated. The center graph depicts the biasing voltage of the The auxiliary,

2,& 49,88 1

11 first thyratron H. The lowest curve represents the square wave output of the multivibrator.

Atthe left-hand end of the curve in Fig. 2(a), the second ignitron I9 is represented as having just become conductive. When the voltag impressed across the third ignitron 2| exceed the voltage impressed across the second ignitron, the third ignitron 2| fires. The first three ignitrons fire in sequence during the time interval represented by the left-hand of curve 2(a). The current input (curve 230) to the welding transformer increases durin this part of the cycle. While the current (curve 230) to the welding transformer 21 is flowing in one direction shown in Fig. 1a by a line above the neutral axis, the output of the multi-vibrator is of one polarity shown in curve as being above its fictional neutral axis 235. When the polarity of the multivibrator output changes, no ignitron of the first three IT, IS or 2| can become newly conductive. The ignitron which is conductive at this time continues conductive even after the voltage through it has become negative. Thi condition arises because the back E. M. F. produced by the decay of flux in the load counteracts th negative potential of the source and the net potential across the last ignitron to conduct continues positive. Thus, the ignitron acts as an inverter during this time, the current output of the welding transformer being supplied to the source and the flux in the transformer decaying to zero.

With the polarity of the multivibrator reversed the fourth, fifth or sixth ignitrons 29, 3| or 33, respectively, may become conductive in succession. One of them becomes conductive only after the last ignitron of the first group l1, l9, 2| becomes non-conductive. This newly conductive ignitron need not be selected by a selecting and timing system. Ignitrons 29, 3| and 33 cannot conduct so long as current flows to the auxiliary anode of one of the ignitrons I1, I9, 2|, that is, so long as one of the latter is conductive. The current output of ignitrons 29, 3|, 33 during this time is shown below the neutral axis 235 of Fig. 2(a) since the last three ignitrons are oppositely connected to th first three ignitrons. During the interval represented by the central of the curve 2(a), the output current to the welding transformer increases in negative value. The last three ignitrons 29, 3| and 33 conduct current alternately until the output of the multivibrator changes polarity. The ignitron conductive as the multivibrator changes polarity acts a an inverter forcing the load current toward zero, and at the proper time the first three ignitrons I, I9 and 2| become conductive alternately for a period of time.

Fig. 2b shows the biasing voltage of the thyratron which controls the conductivity of the first ignitron. The wave (curve 239) of varying amplitude represents the voltage impressed on the firing transformer 224 from the third winding l9] of the fourth secondary 225 of the control transformer I45. The magnitude of this voltage can reach a high maximum 24| only when none of the last three ignitrons 29, 3|, or 33 is firing. The triangular peaks 243 represent the output of the peaking transformer H9 associated with the control circuit of the first ignitron [1. These peaks are impressed only when the multivibrator output is of the one polarity shown in the first of the curve c. When the multivibrator output is of the other polarity the peaking transformers associated with the first three ignitrons emit no pulse. The upper line 245 of Fig. 2(1)) represents the potential level that the grid I of the first thyratron |8| must reach before the first thyratron I81 can fire the first ignitron As can be seen from the drawings, the voltage 244 required to fire the first ignitron is a combination of a peak 243 from the peaking transformer I19 and a high maximum 24| from the firing transformer |9|. The low amplitude portion 25| of the curve 239 represents the condition which arises when one of ignitrons 29, 3| and 33 is conductive. Under such circumstances, the firing transformer in the grid circuit of the firing thyratron associated with the ignitron H is in effect short-circuited and the firing potential is too small to fire the thyratron. Similar curves could be presented for the other ignitron firing circuits.

Although I have shown and described a preferred embodiment of my invention, I realize that many modifications thereof are possible without departing from the spirit and scope of the invention. I do not intend, therefore, to limit my invention to the specific embodiment disclosed.

I claim as my invention:

1. In combination first terminals for deriving direct current, second terminals for supplying alternating current through a load, a first main valve connected in circuit with said first terminals to conduct current therefrom to said second terminals, a second main valve connected in circuit with said first terminals to conduct current therefrom to said second terminals, a control network connected to said second valve to control the conductivity of said second valve, interconnections between said first main valve and said control network for causing said control network to render conductive said second main valve during substantially all the time that said first main valve is non-conductive.

2. In combination first terminals for deriving direct current, second terminals for supplying alternating current through a load, a first main valve connected in circuit with said first terminals to conduct current therefrom to said second terminals, a first control network connected to said first main valve to control the conductivity of said first main valve, a second main valve connected in circuit with said first terminals to conduct current therefrom to said second terminals, a second control network connected to said second valve to control the conductivity of said second valve, a third main valve, a third control network connected to said third valve to control the conductivity of said third main valve, a two-position switch in said third control network which in one position blocks said third main valve from becoming conductive and in the other position allows said third main valve to become conductive.

3. In combination first terminals for deriving electric power; second terminals for supplying power through a load, afirst main valve connected in circuit with said first terminals to conduct current therefrom to said second terminals, a second main valve, a third main valve, connected in cirsuit with said first terminals to conduct current therefrom .to said second terminals, said second main valve being intitially conductive, a first control network having in circuit a first discharge valve which conducts current to said first control network to reach a stationary potential when said second main valve becomes conductive, said first control network being connected to said second main valve and said first main valve and adapted to render said first main valve conductive at the end of a predetermined interval of time after said second main valve becomes nonconductive, a second control network having in circuit a second discharge valve which causes said second control network to reach a stationary potential when said third main valve becomes conductive, said second control network being connected to said third main valve and said second main valve and adapted to render conductive said second main Valve at the end of a predetermined interval of time after said third main valve becomes non-conductive, a third control network connected between said first main valve and said third main valve for causing said third main valve to become conductive at the end of a negligible interval of time after said first main valve becomes non-conductive, and interconnections between each two of the three said main valves so as to prevent the other two said main valves from conducting current when any one said main valve first becomes conductive.

4. In combination terminals for deriving direct current, second terminals for supplying power through a load, a first main valve connected in circuit with said first terminals to conduct current therefrom to said second terminals, a second main valve, a third main valve connected in circuit with said first terminals to conduct current therefrom to said second terminals, said second main valve being initially conductive, a first control network having a two-position switch which in one of said positions prevents said third main valve from becoming conductive but in said second position, connecting said third main valve and said first main valve through a timing circuit, operates to render said third main valve conductive at the end of a negligible interval of time after said first main valve becomes non-con-v ductive, a second control network connected to said third main valve and said second main valve so that at the end of the predetermined interval of time after said third main valve becomes nonconductive said second main valve becomes con ductive, a third control network connected between said second main valve and said first main valve which causes said first main valve to become conductive at the end of a negligible interval of time after said second main valve becomes non-conductive, and interconnections between each two of the three said main valves which prevent the other two said main valves from conducting current when any one said main valve first becomes conductive.

5. In combination first terminals for deriving electric power, second terminals for supplying power through a load, a first main valve connected in circuit with said first terminals to conduct current therefrom to said second terminals, a second main valve, a third main valve connected in circuit with said first terminals to conduct current therefrom to said second terminals, said second main valve being initially conductive, a first control network connected to said second main valve and said first main valve so that at the end of a predetermined interval of time after said second main valve becomes nonconductive said first main valve becomes conductive. a second control network connected to said third. main valve and said second main valve so that at the end of a predetermined interval of time after said third main valve becomes nonconductive said second main valve becomes conductive, a third control network connected between said first main valve and said third main valve having a two-position switch which in one of said positions prevents said third main valve from becoming conductive but in said second position causes said third main valve to become conductive at the end of a negligible interval of time after said first main valve becomes nonconductive, a timing circuit adapted to cause said second main valve to become conductive at the end of a predetermined interval of time after said first main valve becomes conductive, a timing circuit adapted to cause said first main valve to become conductive at the end of a predetermined interval of time after said third main valve becomes conductive and connections between said first, second, and third main valves such that when any one of said main valves becomes conductive it renders non-conductive the main valve then conductive.

6. In combination first terminals for deriving electric power, second terminals for supplying power through a load, a first main valve connected in circuit with said first terminals to conduct current therefrom to said second terminals, a second main valve, a third main valve connected in circuit with said first terminals to conduct current therefrom to said second terminals, said second main valve being initially conductive, a first control network connected to said second main valve and said first main valve so that at the end of a predetermined interval of time after said second main valve becomes non-conductive said first main valve becomes conductive, a second control network connected to said third main valve and said second main valve so that at the end of a predetermined interval of time after said third main valve becomes non-conductive said second main valve becomes conductive, a third control network connected between said first main valve and said third main valve having a two-position switch which in one of said positions prevents said third main valve from becoming conductive but in said second position causes said third main valve to become conductive at the end of a negligible interval of time after said first main valve becomes non-conductive and a capacitor in circuit with the more positive electrode of each two of the three said main valves connected to decrease the voltage difference across the other two said main valves preventing the other two said main valves from conducting current when any one of said main valves first becomes conductive.

7. In combination first terminals for deriving electric power, second terminals for supplying power through a load, a first main valve connected in circuit with said first terminals to conduct current therefrom to said second terminals, a second main valve, a third main valveconnected in circuit with said first terminals to conduct current therefrom to said second terminals, said second main valve being initially conductive, a first control network having in circuit a first discharge valve which causes said first control network to reach a stationary potential when said second main valve becomes conductive, said first control network being connected to said second main valve and said first main valve operating to render said first main valve conductive at the end of a predetermined interval of time after said second main valve becomes non-conductive, a second control network having in circuit a second discharge valve which causes said second control network to reach a stationary potential when said third main valve becomes conductive, said second control networkjbeing connected to said third main valve and said second main valve so that at the end of a predetermined interval of time after said third main valve becomes non-conductive said second main valve becomes conductive, a third control network connected between said first main valve and said third main valve having a two-position switch which in one of said positions prevents said third main valve from becoming conductive but in said second position causes said third main valve to become conductive at the end of a negligible interval of time after said first main valve becomes nonconductive, and a capacitor in circuit with the more positive electrode of each two of the three said main valves which decreases the voltage difference across the other two said main valves preventing the other two said main valves from conducting current when any one of said main valves first becomes conductive.

8. In combination first terminals for deriving electric current, second terminals for supplying alternating current through a load, a first main valve connected in circuit with said first terminals to conduct current therefrom to said second terminals, a second main valve connected in circuit with said first terminals to conduct current therefrom to said second terminals, a control network connected to said second valve to control the conductivity of said second valve, an auxiliary control valve in said control network which conducts current when said first main valve is being rendered conductive so that said control network reaches a non-conductive steady state immediately after said second control network is rendered non-conductive, interconnections between said first main valve and said control network causing said control network to render conductive said second main valve during substantially all the time that said first main valve is non-conductive.

9. Apparatus for controlling the supply of power from an alternating cturent source through a load, comprising in combination a first main valve to be connected in circuit with said source to conduct current therefrom to said load, an auxiliary anode in said first main valve, a second main valve to be connected in circuit with said source to conduct current therefrom to said load, a control network connected to said second valve to control the conductivity of said second valve, and interconnections between said auxiliary anode and said control network caus ing said control network to control the conductivity of said second valve in dependence upon the current conducted by said auxiliary anode.

10. Apparatus for controlling the supply of current from an alternating current source through a load, comprising in combination a first main valve to be connected in circuit with said source to conduct current therefrom to said load, an auxiliary anode in said first main valve, a second main valve to be connected in circuit with said source to conduct current therefrom to said load, a control network connected to said second valve to control the conductivity of said second valve, and interconnections between said auxiliary anode and said control network causing said control network to prevent current from flowing in said second valve so long as current is flowing through said auxiliary anode.

11. Apparatus for controlling the supply of current from an alternating current source through a load, comprising in combination a first main valve to be connected in circuit with said source to conduct current therefrom to said load, an auxiliary anode in said first main valve, a first control network to control the conductivity of said first valve, a second main valve to be connected in circuit with said source to conduct current therefrom to said load, a second control network connected to said second valve to control the conductivity of said second valve, a control system for said networks operating to cause said first network to render said first valve conductive for a predetermined time interval and thereafter to permit said first valve to become non-conductive and connections between said first network and said second network to cause said second network to restrain said second valve from becoming conductive during said predetermined interval and thereafter allows said second network to permit said second valve to become conductive, and interconnections between said auxiliary anode and said control network causing said control network to restrain current from flowing is said second valve so long as current is flowing through said auxiliary anode.

12. Apparatus for controlling the supply of current from an alternating current source through a load, comprising in combination a first main valve to be connected in circuit with said source to conduct current therefrom to said load, an auxiliary anode in said first main valve, a first control network to control the conductivity of said first valve, a second main valve to be connected in circuit with said source to conduct current therefrom to said load, an auxiliary anode in said second main valve, a second control network connected to said second valve to control the conductivity of said second valve, a control system for said networks operating to cause said network to render said valves conductive alternately for predetermined intervals of time, interconnections between said auxiliary anode and said control network causing said control network to restrain current from flowing in said second valve so long as current i flowing through said last named auxiliary anode, and connections between said auxiliary anode of said second valve and said first network for causing said first network to restrain current from flowing in said first valve so long as current is flowing through said last-named auxiliary anode.

13. Apparatus for controlling the supply of current from a polyphase alternating current source through a load, comprising in combination a first plurality of valves to be connected in circuit with said source to conduct current of one polarity therefrom to said load, an auxiliary anode in each valve of said first plurality of valves, a second plurality of valves to be connected in circuit with said source to conduct current of the other polarity therefrom to said load, a control network connected to said second plurality of valves to control the conductivity of said second plurality of valves, and interconnections between said auxiliary anodes and said control network causing said control network to control the conductivity of said second plurality of valves in dependence on current conducted by said auxiliary anodes in any valve of said first plurality of valves.

14. Apparatus for controlling the supply of current from a polyphase alternating current source through a load, comprising in combination a first plurality of valves to be connected in circuit with said source to conduct current of one polarity therefrom to said load, an auxiliary anode in each valve of said first plurality of valves, a second plurality of valves to be connected in circuit with said source to conduct current of the other polarity therefrom to said load, an auxiliary anode in each valve of said second plurality of valves, a control network connected to an auxiliary anode in each valve of said second plurality of valves to control the conductivity of said first plurality of valves, and interconnections between said first auxiliary anodes and said control network causing said control network to control the conductivity of said second plurality of valves in dependence on current conducted by said first auxiliary anodes in any valve of said first plurality of valves.

15. Apparatus for controlling the supply of current from a polyphase alternating current source through a load, comprising in combination a first plurality of valves to be connected in circuit with said source to conduct current of one polarity therefrom to said load, a first auxiliary anode in each valve of said first plurality of valves, a second plurality of valves to be connected in circuit with said source to conduct current of the other polarity therefrom to said load, a second auxiliary anode in each valve of said first plurality of valves, a block network connected to an auxiliary anode in each valve of said second plurality of valves to block the conductivity of said second plurality of valves, and interconnections between said second auxiliary anodes and said control network causing said block network to control the conductivity of said first plurality of valves during the time that current flows to said second auxiliary anodes.

16. Apparatus for controlling the supply of current from an alternating current source through a load, comprising in combination a first plurality of valves to be connected in circuit with said source to conduct current of one polarity therefrom to said load, a first auxiliary anode in each valve of said first plurality of valves, a second plurality of valves to be connected in circuit with said source to conduct current of the other polarity therefrom to said load, an auxiliary anode in each valve of said second plurality of valves, a first transformer having one winding connected in circuit with each of said second auxiliary anodes, so that one winding of said first transformer is shorted when current is flowing to any one of said second auxiliary anodes, a second transformer connected to each one of the first plurality of valves so as to control the conductivity of said first plurality of valves and so as to block conductivity of said first plurality of valves when said second transformers are shorted, said second transformers being connected to one winding of said first transformer so that each of said second transformers is short-circuited when current flow to any one of said second auxiliary anodes shorts the windings of said second transformer.

17. Apparatus for controlling the supply of current from an alternating current source through a load, comprising in combination a first plurality of valves to be connected in circuit with said source to conduct current of one polarity therefrom to said load, a first auxiliary anode in each valve of said first plurality of valves, a second plurality of valves to be connected in circuit with said source to conduct current of the other polarity therefrom to said load, a second auxiliary anode in each valve of said second pluvalves when said second transformer is shorted,

said second transformers being connected to one winding of said first transformer so that each of said second transformers is short-circuited when current flow to any one of said second auxiliary anodes shorts the windings of said second transformer, and connections between said second transformers and each of said first auxiliary anodes so that current fiow to each of said first auxiliary anodes shorts each of said second transformers.

18. Apparatus for controlling the supply of current from an alternating current source through a load, comprising in combination a first plurality of valves to be connected in circuit with said source to conduct current of one polarity therefrom to said load, a first auxiliary anode in each valve of said first plurality of valves, a second plurality of valves to be connected in circuit with said source to conduct current of the other polarity therefrom to said load, a second auxiliary anode in each valve of said second plurality of valves, a first transformer having one winding connected in circuit with each of said second auxiliary anodes so that one windin of said first transformer is shorted when current is flowing to any one of said second auxiliary anodes, a second transformer connected to each one of the first plurality of valves to control the conductivity of said first plurality of valves and to block conductivity of said first plurality of valves when said second transformer is shorted, said second transformers being connected to one winding of said first transformer so that each of said second transformers is short-circuited when current flow to any one of said second auxiliary anodes shorts the windings of said second transformer, connections between said second transformers and each of said first auxiliary anodes so that current flow to each of said first auxiliary anodes shorts each of said second transformers, a first plurality of rectifiers in circuit with the first auxiliary anodes and said second transformer windings so that current flowing to any of the first auxiliary anodes cannot short the second transformer windings, and a second plurality of rectifiers in circuit with the second auxiliary anodes and said second transformer windings so that current flowing in any of the second auxiliary anodes cannot short the first transformer windings.

RICHARD L. LONGINI.

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

UNITED STATES PATENTS Number Name Date 2,356,859 Leathers et a1; Aug. 29, 1944 2,372,964 Livingston Apr. 3, 1945 2,397,089 Cox et al Mar. 26, 1946 2,428,586 Rose Oct. '7, 1947 2,447,133 Nims Aug. 1'7, 1948