US2840762A - Electric discharge apparatus - Google Patents

Description

June 24, 1958 A F. E. MCLANE ELECTRIC DISCHARGE APPARATUS 3 Sheets-Sheet 1 Filed May 7, 1956 RTI DKI

DGI

DAI

Source Control Potential Fig. 20

Fig. 2b.

INVENTOR Fletcher E. Mc Lone.

WITNESSES ATTORNEY Fig.2c.

Fig. 2d.

Fig. 2e.

Fig. 2f.

Fig. 29.

Fig. 2h.

Fig. 2j.

3 Sheets-Sheet 2 ,Kl ,D

MTI Saturated F. E. M LANE ELECTRIC DISCHARGE APPARATUS 4| Conducting 4| Non-Conducting 4i Non-Conducting 4! Conducting MTI Being Converted To Saturation l l" kg Period- June 24, 1958 Filed May 7, 1956 Fig.2k.

June 24, 1958 F. MCLANE I 2,840,762

ELECTRIC- DISCHARGE APPARATUS Filed May 7, 1956 3 Sheets-Sheet 3 B s I I f G01e m I l l) II I H l 1": 2 l I o F I9. 3 "s It 30 Volts Forward Resistance Forward Resistance About 2,000 (1 About 2,000 n llllllllllllllllllllllll mew? or esm ETD Ripple Component F7]- F|g.

ELECTRIC DISCHARGE APPARATUS Fletcher E. McLane, Lancaster, N. Y., assignor to Westinghouse Electric Corporation, East Pittsburgh, 13s., a corporation of Pennsylvania Application May 7, 1956, Serial No. 585,913

Claims. (Cl. 315--168) This invention relates to electric discharge apparatus, and has particular relation to apparatus for controlling the conduction of electric discharge devices, particularly of the gaseous or discontinuous control type.

Where precision in the conduction of an electric discharge device is required, it is essential that the discharge device be controlled by a potential having a steep wave front. This concept was first disclosed by Joseph Slepian in 1923 in an application which matured as Patent 2,327,971. Since that time, there has been a demand for apparatus capable of producing a readily controllable wave potential having a steep wave front for firing a discharge device. It is essential that the position of the steep wave front of the potential be readily shiftable with respect to the potential being supplied between the anode and the cathode of the discharge device and that the precise phase position of the steep .wave front be consistent at any phase position setting. Further, it is essential that the control be of low cost and of relatively small dimensions and have a minimum of moving parts subject to wear.

The use of so-called Ramey magnetic amplifiers for obtaining the steep wave front control has been suggested. The Ramey magnetic amplifier is a highly saturable transformer having a gating winding and a reset winding. When the Ramey amplifier is applied to control a discharge device, the gating winding and a gating source of sinusoidal potential are connected between the control electrode and the cathode, and the reset winding is supplied with reset potential which may vary in magniture. The gatingpotential leaves the transformer saturated at one polarity and the magnitude of the reset potential impressed at any time determines the extent of magnetization of the transformer at the beginning of the subsequent gating interval.

Thus for resetting the equation applies, where e,- is the applied reset voltage, N the number of turns of the reset winding, o the flux at the end of resetting and g5, the flux at the beginning of resetting. The resetting potential leaves the flux at a magnitude which may vary from a magnitude of intermediate magnetization to saturation of the opposite polarity (here The resetting potential may be direct or alternating current potential.

The gating potential drives the transformer to saturation during the gating interval in a subinter-val which is derivable from the equation where e is the impressed gating potential, 5 is the magnetic flux, o being the flux corresponding to initial magnetization and the flux at saturation, N is the number of turns of the gating Winding and t is the time. At the time t when the transformer reaches saturation, the peters nited States Patent 2,840,762 Patented June 24, 1958 tial absorbed by the transformer drops abruptly to a low magnitude and the source potential is applied between the control electrode and the cathode of the discharge device to control its conductivity.

Where 6 is sinusoidal the above equation becomes where s is the amplitude of the gating potential and p its frequency. The amplitude s is such that when the reset potential saturates the transformer the time t is of the order of a half period. Under such circumstances the lower limit p, is the above equation then becomes or for other magnitudes of t 4 s and thus the instant when the potential between the control electrodes and the cathode rises abruptly to the magnitude of the impressed gating voltage is at a phase position dependent on In attempting to apply the Ramey magnetic amplifier as just disclosed, several disadvantages have been met. Among these, the most important is that with the abovedescribed control circuit, precise control over the complete half period of the anode-cathode potential is not achieved. Another disadvantage is that the control in the region approaching conduction near the end of the half period; that is, conduction of a low magnitude, is not precise because in this region the steep wave front control potential is of relatively small amplitude.

It is, accordingly, broadly an object of this invention to provide a relatively low-cost control circuit of small dimensions for precisely controlling the conductivity of an electric discharge device.

Another object of this invention is to provide such a control circuit with which it shall be feasible to impress abrupt potential to control the discharge device over the full half period of anode-cathode potential impressed thereon and the magnitude of the control potential abruptly impressed shall be substantial throughout the full half period.

A more specific object of this invention is to provide a control circuit for an electric discharge device including a magnetic amplifier transformer of the Ramey type which shall operate to control the discharge device over the full half period of anode-cathode potential impressed thereon, and which shall provide a substantial abrupt control potential over the full half period.

In accordance with this invention, a control circuit for an electric discharge device is provided which includes in circuit with the control electrode and cathode of the discharge device the gating winding of a magnetic-amplL fier transformer and a gating source that impresses on the gating winding a potential of fiat top wave form. Specit 8 6K] dt=N 02 0 where-e is-theconstant magnitude of the gating potential. That-is-e t=N(,,) and T6 ZNQSQ. re

Thus, with the fiat-top wave form potential of the proper magnitude applied and the transformer partially reset, the time required for the transformer to revert to saturation from the partial resetting magnetization depends on theextent of thismagnetization and is in duration less than the duration of the gating half period by a magnitude depending on this magnetization. At the instant when a partially reset transformer reaches saturation, the potential in the control circuit of the discharge device varies abruptly, and this abruptly varying potential is impressed to control the conductivity of the discharge device. By applying a resetting potential of the proper magnitude, the abrupt change maybe produced at any instant in the gating half period over the full half period of supply potential, and since the potential is of flat top throughout the half period, the abrupt change in potential is the same throughout the half period.

Thenovel features considered characteristic of this invention are described generallya'bove. The invention itself both as to its organization and its method of operation together with additional objects and advantages thereof will be understood from the following description of a specific embodiment taken in connection with the accompanying drawings, in which:

Figure 1' is a circuit' diagram showing a preferred embodiment of this invention;

Figs; 2a through 2k are graphs illustrating the operation of Fig. 1;

Fig. 3 is a graph showing the magnetization curve of a magnetic amplifier transformer used in the practice of this-invention; and

Fig. 4 is a circuit diagram showing the dimensions of thecomponents used in an embodiment of this invention which has been tested and found to operate satisfactorily.

Fig. 4 is presented in an effort to help those skilled in the art to practice this invention and without any intention of in any way limiting the scope of this invention. This invention may be practiced with components differingfrom those disclosed in Fig. 4 both as to kind and as tomagnitude, and apparatus similar to Fig. 4, but including such different components, are within the scope of this invention.

Theapparatus shown in Fig. 1 comprises a load supplied through a pair of thyratrons T1 and T2, each including an anode A1 and A2, a cathode K1 andKZ and a control electrode G1 and G2. Power is supplied to the load from conductors L1 and L2, which may be energized from a commercial supply, through the primary P of a transformer T having a secondary S with an intermediate tap 11. The load is connected between the in- 4 termediate tap 11. andacommonjunctionof the cathodes K1 and K2 of the thyratrons T1 and T2. The anodes A1 and A2 are respectively connected to the terminals of the secondary S. The circuit including the load and the transformer T is typical of the thyratron supply circuits to which this invention is applicable.

The control potential impressed on each of the thyratrons T1 and T2 is composite including a blocking biasing component and a counteracting component. The blocking biasing component is derived from a biasing source 13 common to the thyratrons. The positive terminal of the source is connected to the junction of the cathodes K1 and K2, and the negative terminal is connected to each of the control electrodes through resistors 15 and 17, and grid resistors 16 and 18, respectively. The counteracting component in the control circuit of each of the thyratrons T1 and T2 includes a direct current source 21, asource of half period pulses 23 and 25, respectively, and a magnetic amplifier transformer MT1 and MT2, re.- spectively, of the. highly saturable type. The source 21 is common to the thyratrons. Each of the transformers MT1 and MT2. includes a gating windingGTl and GT2 and a reset winding RT1 and RT2. The half period pulses for the control circuits of the thyratrons are derived from the conductors L1 and L2 through an auxiliary transformer AT1, having a primary APl and secondaries 1AS1. and 2AS1. The secondary lASl has an intermedi ate tap 31. Each section of the secondary 1AS1 is connected across a resistor. 33 and 35, respectively, through a rectifier 37 and 39, respectively. When the transformer AT1 is energized, half period pulses appear across the resistors 33 and 35 during succeeding alternate half periods of the supply.

The direct current potential source 21 is connected at its negative terminal to the negative terminal of the blocl ing bias 13, and at its positive terminal to the junction of the resistors 33 and 35. Each of the resistors 33 and 35 is connected at its remaining terminal to one terminal of the gating winding GT1 and GT2, respectively, of the associated magnetic amplifier transformer. The other terminal of each of the gating windings GT1 and GT2 is connected through a rectifier 41 and 43- to the control electrode G1 and G2, respectively, of the associated thyratron.

The control potentials for thyratrons T1 and T2 thus each consist of a blocking bias 13 which is modified by reaction across resistors 15 and 17 by the algebraic sum of the potentials appearing across winding GT1 and resistor 33 and the potential 21 and winding GT2 and resistor 35 and potential 22. In situations in which the characteristics of the thyratron are such that they require a substantial positive potential to be rendered conducting, the blocking bias (13) may be omitted. The potentials across GT1 and GT2 are determined by the reset potential impressed across RT1 and RT2.

The reset windings RT1 and RT2 are each supplied with reset potential through an associated section of a double triode DT. Each section of the double triode has an anode DA1 and DA2, a cathode DKl and DKZ and a control electrode DGl and DG2. The reset potential is derived from the secondary 2AS1 which has an intermediate tap 51. The intermediate tap 51 is connected through a blocking biasing potential 53 to the common junction of the cathodes DKl and DKZ. The terminals of the secondary are connected each through a reset winding RT1 and RT2 to the anodes DA1 and DA2, respectively. The control electrodes D61 and DG2 are connected in common to a source of control potential which determines the conductivity of the sections and thus the magnitude of the current conducted through the reset windings RT1 and RT2. The potential of the secondary 2AS1 connected to the reset windings RT1 and RT2 is so phased with respect to the potential of the secondary 1AS1 connected to the gatingwindings GT1 and GT2 that the reset potential is supplied in opposite phase to the gating potential to each of the magnetic amplifier transformers MT1 and MTZ.

In the operation of the apparatus, the source of control potential is set at the desired magnitude so that the sections DA1-DK1 and DA2-DK2 have the desired conductivity. Reset current of preset magnitude is then supplied during alternate half periods of the supply to the windings RT1 and RT2. For apparatus such as is shown in Fig. l, the reset currents flowing through each of the windings RT1 and RT2 should preferably be of substantially the same magnitude and, if necessary, balancing provisions to achieve this purpose may be included in the anode and control circuits of the sections DAl-DK1 and DA2-DK2. Conversely, under certain circumstances, the balancing may' be achieved by supplying resetting currents of different magnitudes. The secondary S should be so poled relative to the secondary 2AS1 that the reset current is supplied to RT1 during the half periods when the anode A1 is negative relative to the cathode K1 and through RTZ during the half periods when the anode A2 is negative relative to the cathode K2. During the intervening half periods in each case, when the anode Al is positive relative to the cathode K1 and the anode A2 is positive relative to K2, gating potential is impressed on GT1 and GT2, respectively. Transformers MT1 and MT2 are saturated at instants in the latter half periods depending on the magnitudes of the reset currents. At these instants the impressed potential is in efiect transferred from GT1 and GT2 to resistors 15 and 17, respectively, and the associated thyratrons T1 and T2 are rendered conducting. The cooperation of the transformers MT1 and MT2, the pulse potential sources (33 and 35) and the direct-current source 21 are such that the conductivities of thyratrons T1 and T2 may be varied precisely over the full half periods of positive anode-cathode potential impressed on them by varying the magnitude of currents conducted by the sections DAlDKl and DAZ-DKZ. The manner in which this precise control is achieved will now be explained with reference to Figs. 2 and 3. Since the sections of Fig. 1, including the thyratrons T1 and T2 operate similarly, the explanation will be presented with reference to one of the sections, the one including T1, and Fig. 2 refers only to this section.

Fig. 2 includes a number of graphs (a through k), in

each of which potential is plotted vertically and time horizontally.

The time coordinates for all the graphs a through k which are aligned vertically are assumed to occur at the same instant. in each case, the potential plotted vertically is referred to a labeled point in the selected portion of Fig. l, and where necessary, Fig. l is labeled with a capital letter to correspond to Fig. 2.

In graph a, the potential impressed between the anode A1 and the cathode K1 of the thyratron T1 is plotted as a function of time; this potential appears between A1 and K1 when T1 is non-conducting and is of ordinary sine wave form. In graph 1;, the half wave potential between the points B and C, that is, across the resistor 33, is plotted. It is seen that this half wave potential is in phase with the potential between the anode A1 and the cathode K1. Graph c is a plot of the potential of point D, the negative terminal of the direct-current supply 21 relative to point B. This graph is a straight line below the B axis, a distance corresponding to the magnitude of the supply 21.

Graph :1 is a plot of the potential of point C relative to D and is derived by superimposing the 'curve of graph b on the line D of graph c. It is seen that the wave form of curve of graph d consists of a sine wave loop preceded and followed by a trapezoidal loop. The base of the trapezoid is set by the magnitude of the potential 21 (Fig. 1) and may, for example, be of a duration of the order of of a period, that is, approximately 200.

The fiat top of the trapezoid has a duration of the order of one half period of the supply.

Graph e presents the potential of point D relative to the cathode K1 as a function of time. Since this potential is negative by the magnitude of the blocking bias 13 (Fig. l), the curve of graph 2 is a straight line. Graph f represents the potential of the point C relative to the cathode K1, and this curve is derived by superimposing the curve of graph d on the line D of graph e. The wave form of this curve is similar to that of graph d, but the height of the top of the trapezoid above the abscissa is smaller than in the case of graph d by the magnitude of the bias 13 (Fig. l).

The potential represented by graph is impressed on the control electrode G1 of thyratron T1 through the gating winding GT1 of the magnetic amplifier transformer MT1 and through the rectifier 41, and is modified by the impedance of the transformer MT1. The manner in which the transformer MT1 performs its modifying function may be understood with the help of Fig. 3 which is a magnetization curve for the transformer MT1.

In Fig. 3, the magnetic induction B is plotted vertically as a function of the field H which is proportional to the current supplied through sections DA1-DK1. The

square line curve represents the hysteresis loopfor themagnetic transformer MT1. It is seen that while the horizontal boundaries of this curve are substantially rectilinear, the vertical boundaries have a slope and it is because of this slope that the use of the magnetic transformer MT1 for controlling the phase of the potential impressed on the control electrode G1 is feasible. The gating and resetting of the transformer MT1 is effected by impressing potentials through the gating and resetting windings GT1 and RT1, respectively, during succeeding half periods. The resetting potential has been discussed; the gating potential is derived from the points D and C and has the form shown in Fig. 2d. The effect of the gating potential transmitted during any half period is determined by the magnetic condition of the transformer -MT1 as a result of the preceding resetting excitation during the preceding half period. Assume that the transformer MT1 is at any instant in a saturated condition, the polarity of the saturation corre- 'sponding to the point I (fluX+,) of Fig. 3 and that current suflicient to revert the transformer MT1 to saturation of the opposite polarity is transmitted during a resetting half period. This current should produce a field having a magnitude hl. The magnetization of the transformer MT1 then varies from I to point II along the left hand portion of the magnetization curve as indicated by the heavy arrows in Fig. 3, and at the end of the resetting half period, the transformer MT1 is in the condition represented by point II (fluxb The gating of the transformer MT1 may now be considered with reference to graph d of Fig. 2.

Assume that the resetting half period just mentioned corresponds to the first half period on the left of Fig. 2d. During this half period, the potential impressed at point C is negative with respect to the potential impressed at point D, and conduction through the gating Winding GT1 is blocked by the rectifier 41. The gating starts at the instant when the potential of point C becomes positive relative to point D; that is, at the left hand terminal of the base of the first trapezoid of Fig. 2d, and continues until the potential of point C again becomes negative. The magnitude of the direct current potential 21 and of the half wave potential (33) impressed between the points and C is such as to produce a field h2. This potential applied to a transformer in the condition represented by the point II of Fig. 3 reverts the transformer to the saturated condition represented by the point I 5,) of Fig. 3. The magnetization of the transformer during the reversion effected during the gating interval varies along the right hand portion of the loop in Fig; 3 as indicated by the heavy arrows. During this reversion, the impedance of thetransformer MTi is high and the drop produced by the gatingpotential is substantially absorbed by the transformer.

The above described operationis repeated during succeeding resetting and gating half periods. The potential of the control electrode G1 is in this situation not materially affected by the potential at C. during the gating interval since the potential is largely absorbed by the gating winding GT1. This situation is represented in Fig. 2g in which. the potential of the control electrode G1 with reference to the cathode K1 is plotted. This curve consists of a series of upwardly and downwardly extending lines. The lowermostlines correspond to the half periods in which the rectifier 41 is non-conducting andthepotential of the control electrode G1 relative to the cathode K1 is equal to theblocking bias potential 13. During the half periods during which'the rectifier 41 is conducting, a large proportion of the potential between the point C and the point D is absorbed by the gating winding GT1 and the small remaining potential is represented by the uppermost. lines. It is seen that with the double triode controlled-so that the transformer MTl is fully reset, the potential of the control electrode G1 relative to the cathode K1. remains substantially negative; the thyratron' T1 is then non-conducting.

Nov/consider the situation in which the potential'z between the control electrode D61 and the cathode DKl of the section DAll-DY1 of. the double triode associated with the transformer MTl' is so set that resetting current through this section produces substantially zero resetting. Under such circumstances, the condition of the transformer at the beginning of each gating cycle is represented by the point i of Fig. 3, and the impressing of the gating potential has no effect on themagnetization of the transformer MTl. The potential which is impressed. between the control electrode G1 and the cathode K1 under conditions of zero resetting is represented by Fig. 2h which presents the potential of the control electrode- G1 relative to the cathode K1. The resulting curve consists of alternate. rectangular and trapezoidal loops and has alternate negative and positive fiat. tops. The negative tops correspond to the condition during which the polarity of the potential impressed between the terminals C and D is such that the rectifier 41 is non-conducting, and the potential of the control electrode G1 isequal to the potential at point D. The trapezoidal loops correspond to the intervals during which the rectifier 41 is conducting, and the potential of the control electrode G1 is substantially the potential of. the point C since the impedances of the gating winding GT1 and the rectifier 41 are small. Under such circumstances, the potential of the control electrode G1 is positive during of a period within which is included'the half period of positive anode-cathode potential of thyratron T1. Thyratron Tl is then fully conducting as represented by Fig. 21' in which the potential-of the anode A1 with reference to-the cathode K1- is again plotted. The conducting intervals of the thyratronTl are represented by the half waves in broken lines under which there is shading.

Now consider the situation when thecontrol'potential of section DA1-DK1 is so setthat the transformer MT is only partially reset. Partial resetting maybe effected by setting the current flowing through the resetting Winding at a magnitude such that the field produced lies at a magnitude k3 (p between the fnll resetting magnitude M and a minimum resetting magnitude I13. For partial resetting, the magnetization is partially reduced from the magnitude corresponding to the point I tothe magnitude corresponding to the point Iii. The variation of the magnetization during the partial resetting is represented by the loop of Fig. 3 indentified by the light arrows. It is seen that when the transformer MTlvis partially reset, it is in condition represented bypoint III. Since. the magnetization of the transformer is not at the full saturation. magnitude represented by point II of Fig. 3', the integral of the potential of the voltage by the time impressed. on the gating winding GT1 which is required to revert the transformer to the saturation condition represented by point I is substantially smaller than when the transformer is fullyreset. The transformer MTI then reaches saturation at a point intermediate the beginning and the end of the gating half period.

This condition is represented in Fig. 2 in which the potential of the control electrode G1 with reference to the cathode B11 is plotted. In this case the curve corresponding to each cycle of potential is stepping. During the earliest part of the period, the rectifier 41' is nonconductingand the potential of the control electrode G1 ual to the potential of the negative terminal of the rig supply 13. During the part represented by the next portion of the curve, the rectifier U1 is. conducting and the, transformer MTl is being reverted to saturation.

Therpotential ofthe control electrode G1 is still negative but less negative than during the first stage by the magnitude represented by the large-drop across the gating windingGTl. During the next stage, the transformer MT1 is saturated, and the potential of the control electrode G1 rises substantially to the potential of point C.

The thyratron. T1 is rendered conducting at the instant when the potential rises to the positive magnitude of the point C. The conduction of the thyratron is illustrated in Fig. 2k in which the potential of the anode A1 with In this case the.

reference to the cathode K1 is plotted. shaded areas under the broken line sine wave. loops represent theconduction, and it is seen that the conduction:

starts at a predetermined angle in each of the half periods.

It is to be noted that the time of response of the novelapparatus disclosed herein is one half period of the al ternating potential. That is a variation in the resetting potential impressed at the beginning of one half period produces a response during the next half period.

While a preferred embodiment of this invention has been disclosed herein, many modifications thereof are feasible. The invention, therefore, is not to be restricted except insofar as is necessitated by the spirit of the prior art.

I claim as my invention:

1. In combination an electric discharge device having an anode, a cathode and a control electrode, a magneticamplifier transformer of the highly saturable type having a gating winding and a reset winding, means connected between said anode and cathode for impressing an alternating potential between said anode and cathode, means connecting said gating winding in circuit with said control electrode and cathode, said connecting means including means for impressing in said circuit, through said gating;

winding, gating potential of fiat-top wave form during alternate half periods of said alternating potential, means connected to said reset winding for impressing on said reset winding a reset potential effective during the intervening alternate half periods of said alternating potential, and means for controlling the magnitude of said reset potential, said fiat-top of said gating potential having a duration of substantially one-half period of said alternating potential and a magnitude such that if said transformer is reset to saturation at one polarity at the beginning of a gating half period and said gating potential is impressed during the whole of said half period across said gating winding said transformer passes during the gating half period to saturation of the opposite polarity. V

2. In combination an electric discharge device of the discontinuous control type having an anode, a cathode and a control electrode, means connected between said anode and cathode for impressing an alternating potential between said anode and cathode, a magnetic-amplifier transformer the highly saturable type including a gating winding and a reset winding, potential absorbing means, means connected to said potential absorbing means.

9 Wave potential pulses during alternate half periods said alternating potential, direct current potential supply means, rectifier means, means connecting in a circuit in series said control electrode, said cathode, said direct current potential supply means, said potential absorbing means, said gating winding and said rectifier means so as to impress gating potential in said circuit during alternate half'periods of said alternating potential, means connected to said reset winding for impressing reset potential on said reset winding effective during the intervening alternating half periods of said potential, and means for determining the magnitude of said reset potential, said gating potential having a duration substantially of a half period of said alternating potential and a magnitude such that when said transformer is saturated at one polarity by said reset potential at the beginning of a gating half period and said gating potential is impressed on said gating winding during the whole of said gating half period, the magnetization of said transformer is reversed during said last-mentioned gating half period -means, means connected to said potential absorbing means for impressing across said potential absorbing means half wave potential pulses during alternate half periods of said alternating potential, direct current potential supply means, rectifier means, means connecting in a circuit in series said control electrode, said cathode, said direct current potential supply means, said potential absorbing means, said gating winding and said rectifier means so as to impress gating potential in said circuit during alternate half periods of said alternating potential, the potential impressed across the portion of said last-named circuit including said direct current potential supply and said potential absorbing means having a wave form including a sinusoidal loop followed by a trapezoidal loop, the trapezoidal loop having a base of duration of the order of five-ninths of a period of said alternating potential, means connected to said reset winding for impressing reset potential on said reset winding effective during the intervening alternating half periods of said potential, and means for determining the magnitude of said reset potential, said gating potential having a duration substantially of a half period of said alternating potential and a magnitude such that when said transformer is saturated at one polarity by said reset potential at the beginning of a gating half period and said gating potential is impressed on said gating Winding during the whole of said gating half period, the magnetization of said transformer is reversed during said last-mentioned gating half period so that said transformer is saturated at the opposite polarity substantially at the end of said last-mentioned gating half period.

4. In combination an electric discharge device of the discontinuous control type having an anode, a cathode and a control electrode, means connected between said anode and cathode for impressing an alternating potential between said anode and cathode, a magnetic-amplifier transformer of the highly saturable type including a gating Winding and a reset Winding, potential absorbing means connected to said potential absorbing means for impressing across said potential absorbing means halfwave potential pulses during alternate half periods of said alternating potential, direct-current potential supply means, rectifier means, means connecting in a circuit in series said control electrode, said cathode, said direct current potential supply means, said potential absorbing means, said gating Winding and said rectifier means so as to impress gating potential in said circuit during alternate half periods of said alternating potential, the potential impressed across the portion of said last-named circuit including said direct-current potential supply and said potential absorbing means having a Wave form including a sinusoidal loop followed by a trapezoidal loop, the trapezoidal loop having a base of duration greater than onehalf period of said alternating potential, means connected to said reset winding for impressing reset potential on said reset Winding effective during the intervening alternating half periods of said potential, and means for determining the magnitude of said reset potential, said gating potential having a duration substantially of a half period of said alternating potential and a magnitude such that when said transformer is saturated at one polarity by said reset potential at the beginning of a gating half period and said gating potential is impressed on said gating winding during the Whole of said gating half period, the magnetization of said transformer is reversed during said last-mentioned gating half period so that said transfgormer is saturated at the opposite polarity substantially at the end of said lastmentioned gating half period.

5. In combination a saturable transformer having a gating winding and a reset winding, means connected to said reset winding for impressing a reset potential thereon, means for supplying a periodic potential having a Wave form consisting of a half-Wave sinusoidal loop followed by a trapezoidal loop, a rectifier, means connecting in a series circuit said periodic potential supply means, said gating winding and said rectifier so that said periodic potential is impressed in said series circuit, said rectifier being poled to block the potential of said sinusoidal loop and to permit current to flow while said trapezoidal loop potential is impressed, said trapezoidal loop potential having a duration of at least one-half the period of said periodic potential and a magnitude such that when said transformer is saturated at one polarity by said reset potential at the beginning of a gating half period during which said trapezoidal-loop potential is impressed the magnetization of said transformer is reversed to saturation of the opposite polarity during said gating half period.

References Cited in the file of this patent UNITED STATES PATENTS 1,960,047 Bedford May 22, 1934 2,264,695 Gulliksen Dec. 2, 1941 2,366,561 Schmidt Jan. 21, 1945 2,370,287 Bivens Feb. 27, 1945 2,489,858 Burnett Nov. 29, 1949 2,730,661 Kellogg Jan. 10, 1956

Images