US2027383A - Electric discharge device - Google Patents

Description

. i 1936- c. E. KENTY QGZKS ELECTRIC DISCHARGE DEVICE Filed 901;. 3, 1934 ZSheets-Sheet l jam 1% W360 c. E. KENW ELECTRIC DISCHARGE DEVICE Filed Oct. 3, 1934 mzmm 2 Sheets-Sheet 2 ;/CENTER OF ARC TUBE RESULTANT Patented Jan. 14, 1936 2,027,383 ELECTRIC DISCHARGE DEVICE Carl E. Kenty, Chatham, N. J., assignor to General Electric Vapor Lamp Company, Hoboken, N. J., a corporation of New Jersey Application October 3, 1934, Serial No. 746,726

9 Claims.

The present invention relates to electric gaseous discharge devices generally, and more particularly to discharge devices of the type which are operated with a relatively constricted arc.

A particular object of the invention is to provide a novel method and means for controlling the arc path within a discharge device in order to prevent the are from impinging on the envelope wall. A further object of my invention is to produce a magnetic field having a rapidly decreasing intensity within such a device. Another object of my invention is to lessen the variation in arc position during the starting period of the mercury vapor lamp, wherein there is a con- I stricted arc discharge, difliculties are encountered whenever an attempt is made to operate the device with the arc extending in other than a vertical line. These difficulties are caused by the fact that the arc tends to rise, due to the heating of the gas in the arc stream, and thus curves toward and impinges upon the upper wall of the arc tube as the latter is moved toward a horizontal position. The glass which is commonly used I for these arc tubes is not adapted to withstand the ensuing severe localized heating and hence softens, allowing the tube to lose its shape and at the same time evolving copious amounts of gas into the discharge path, either of these efiects alone being enough to ruin the discharge device. lit has been proposed heretofore to overcome this dimculty by the use of a magnetic force which opposes this tendency toward upward movement of the arc stream, and in practice it has been found that this gives very desirable results. A lesser difiiculty still remains however. Thus during the starting period of one of these high intensity mercury vapor arc lamps the arc current is quite high, of the order of double that which flows after the vapor pressure has reached equilibrium, and hence the series magnets which are necessarily used exert a downward force on the arc which is of the order of four times that which is exerted during normal operation of the lamp. 5 At the same time the temperature of the arc stream is still less than that attained during operation, so that the gravitational ,forces which tend to force the arc upwardly are much less than those which'the magnetic force is designed to 10 neutralize. As a result it is found that a magnetic field which is entirely satisfactory during continued operation of a lamp will force the arc downwardly onto the lower wall of the arc tube during starting, failures occasionally occurring as the result of the abnormal heating to which the envelope wall is thus subjected.

I have now discovered that this difficulty is entirely overcome by the use of a novel magnet structure,-which provides a field of rapidly decreasing intensity within the lamp, this field in the preferred form actually reversing before the bottom of the arc tube is reached. Thus only a slight variation in the position of the arc is necessary to completely overcome the effects of the larger currents which flow'during the starting period, while the reverse field makes it impossible for the are ever to reach thepbottom of the arc tube under any conditions.

Furthermore this novel structure provides another novel result. For some reason which is not entirely understood the ends of the arc, adjacent to the electrodes, are forced upwardly more strongly than the middle of the arc, with the result that in any uniform magnetic field the arc assumes a somewhat bowed shape. In'certain types of reflectors this is undesirable since it reduces the efiiciency of the reflector, due to the deviation of the are from a straight line source. With my novel structure, however, the field 4w strength falls ofi so rapidly that this bowingof the arc is so minimized that it is no longer noticeable.

Where it is desired to still further limit the variation in position of the arc during the starting period I have found that the use of a series transformer with a, saturable core very materially aids in reducing the magnetic force during the continued flow or the high starting current. I

I have likewise discovered that where it is desired to produce an absolutely straight are this can be done by a novel structure whereby the magnetic field along the arc path is rendered non-uniform.

All of these novel results are obtained through the use of series coils which materially assist in the necessary stabilization of the are; so that no additional energy is required therefor. An additional advantage of my novel structure is that itreadily lends itself to inclusion within a reflector which is mounted close to said lamp.

For the purpose of illustrating my invention I have shown in the accompanying drawings a preferred embodiment, together with several modifications thereof. In these drawings Fig. 1 is an elevational view of a complete lamp unit embodying my invention,

Fig. 2 is a sectional view of this unit taken on the line 22 of Fig. 1, v

Figs. 3 and 4 are similar sectional views showing modifications of the structure of Figs. 1 and 2,

Fig. 5 is a plan view of the iron core shown in Figs. 2, 3 and 4,

Fig. 6 is a schematic circuit diagram showing the electrical connections employed in the structure shown in Figs. 1 and 2,

Fig. 7 is a graph which shows the variation in the field strength at difierent distances from the reflector,

Fig. 8 is a schematic diagram showing a modification of the electrical connections of Fig. 6,

and

Fig. 9 is a schematic diagram showing another.

modification of my invention.

As particularly shown in Figs. 1, 2 and 6 my novel lamp unit consists of a high intensity mercury vapor arc lamp I of the general type disclosed, for example, in the co-pending application of James A. St. Louis, Ser. No. 724,586, filed May 8, 1934, which is screwed horizontally into a suitable socket (not shown) within the housing 2, together with the V-shaped reflector 3 which is supported by said housing with the angle in said reflector extending along said lamp, and virtually in contact therewith. This reflector, which is made of brass, aluminum, or other non-magnetic sheet metal, is designed to reflect the light from the lamp I in a horizontal direction or slightly below, the unit thus being especially suited for street lighting and other uses where an extremely broad distribution of light is required. A magnet core 4, which is conveniently formed of inch sheet iron or any other suitable magnetic material bent into the form of a channel with converging sides, the gap between these sides being of the order of 2 or 3 millimeters, is positioned in the angle of the reflector 3 with the air gap therein as near the bottom of said reflector as it will go, and extends longitudinally to a point somewhat beyond each end of the inner arc tube 5 of the lamp I, as indicated by the dotted lines in Fig. l. A coil of wire 6 is wound longitudinally on the core 4, fifty to sixty turns ordinarily being employed. Above this magnet there is mounted on suitable nonmagnetic brackets a second one, consisting of the flat sheet iron core I and the coil 8, of the order of fifty to sixty turns, which is longitudinally wound thereon. As shown in Fig. 6 the coils 6 and 8 are connected in series with the arc in the lamp I, the connections being such that the current flows in the coil 6 above the core 4 and in the coil 8 below the core I in the same direction as in the arc itself, as graphically illustrated by the arrows in Fig. 6, which represent an instantaneous condition. An inductance 9, which cooperates with the coils 6 and 8 in stabilizing the operation of the arc, is enclosed within the housing}. Lead wires I also extend into said housing from any suitable source, preferably 220 volts A. C.

With the structure which has been described as soon as potential is applied to the leads I0 an arc is immediately initiated within the arc tube 5, this arc current flowing in series through the coils 6 and 8. At the moment of starting, this current is of the order of double the normal operating current of the lamp I, due to the low vapor pressure of the mercury within the arc tube 5. As a result the magnetic fields produced by the coils 6 and 8 are likewise doubled, and the magnetic forces exerted thereby on the arc stream are therefore quadrupled. The fields produced by these two coils are opposed to each other, however, and due to the fact that they are at different distances from the arc path and also due to the radically different arrangement of the cores 4 and I, these two fields fall off at entirely different rates in the region occupied by the arc tube 5. Thus as clearly shown in Fig. 7 the field due to the coil 6 is falling off extremely rapidly with increases in the distance from the apex of the reflector 2, the core 4 being so shaped as to accentuate this characteristic, and also located as close to the lamp I as possible, in order toutilize the steepest portion of the field gradient. The field due to the coil 8 falls oif much more slowly, the core I being kept at such a distance and being so shaped as to make this field as near constant as possible in the area occupied by the arc tube 5. The relative intensities of these fields are so chosen that they exactly neutralize each other at a point only slightly below the center of the arc tube 5, the point at which this neutralization occurs obviously being independent of the current flowing through the coils 6 and 8. As a result, therefore, of this novel combination it is obvious that even though the individual magnetic reaction between the arc stream and each oi the separate magnetic fields is quadrupled, as noted hereinbefore, there is a point slightly below the center where these reactions are completely neutralized, so that it is impossible for the fields to depress the arc stream below this point in the tube. Hence it is impossible for the arc to come into contact with the bottom of the arc tube during the starting period, regardless of how small are the gravitational forces which oppose the magnetic downward thrust, or of how large the starting current is. Furthermore, if the arc stream should by an chance wander below this neutral field position it is obvious that the reversed field would of itself tend to force it back toward the the middle 01' the arc tube.

As the lamp I warms up the vapor pressure therein rapidly increases, as is well known, and

the arc current decreases until finally equilibrium is attained. At the same time the temperature of the gas in the arc stream increases, and this results in an increased tendency for the arc stream to rise. As a result the arc stream gradually moves upwardly within the arc tube 5 from its original position near the neutral field point toward the center of the tube, until equilibrium is attained with the are operating exactly at the center of the tube, or slightly below, as desired. The steepness of the field gradient which results from my novel magnet structure is or great assistance in holding the are at the desired central 'point, since any upward deviation of the are results in a vast increase in the magnetic force exerted, while a downward deviation results in a correspondingly decreased force, making the position of the are quite critical. Likewise this steep gradient permits the arc to be operated at the start only a slight distance below the normal operating position, as hereinbefore noted. This steep gradient is also of advantage in another respect, however. The are in the arc tube has an unequal tendency toward upward movement at different points along-the arc. Thus in the vicinity of the electrodes this force is greater than at the middle of the arc. As a result the arc tends to bow upwardly near the ends and downwardly near the middle when a magnetic force is applied thereto to keepit at the center of the are tube. With mynovel structure this bending of the arc is almost completely eliminated, due to the fact that the rapid changes in field strength with slight changes in position minimize the amount of bowing necessary for all parts of the arc stream to be in equilibrium, with the result that the are in my novel outfit appears to be practically straight.

Even the slight remaining tendency toward bowing of the arc stream'is eliminated, however, with the structure illustrated in Fig. 5, where the horizontal portion of the core 4 is shown as having a; plurality of perforations it near the middle thereof, with the result that the field at the middle of the arc is made slightly, say weaker than that in the vicinity of the electrodes. It is also to be observed that for best results the gauze band I2 on the arc tube 5. which serves as a starting electrode, should be made of a non-magnetic material, so as to avoid any magnetic shielding of the enclosed portion of the arc.

Somewhat similar results are obtained with the modified magnet structures shown in Figs. 3 and 4. In each of these structures only the single coil 6, of thirty or forty turns, wound on the core t is employed, this coil being connected in series with the arc in the same manner as shown in Fig. 6. In Fig. 3 anadditional core i3, of similar shape to the core 4, is placed above the coil 6 with the edges thereof extending downwardly and overlapping the sides of the core 4, but with an air gap of the order of a few millimeters therebetween. This magnet arrangement has been found to give a very steep field gradient in the area occupied by the arc tube 5,

so that there is very good control of the arc path. This field, however, does not fall off 'to zero, as is true with the preferred structure of Fig. 2, and hence the arc is depressed further during the starting period, although it does not come into contact with the bottom of the arc tube 5 at any time. In Fig. 4 only the single core l is employed, and this simplified structure has likewise been found to give a rapidly falling field,

as graphically illustrated in Fig. '7, so that this arrangement is also entirely satisfactory under some conditions, as where a somewhat greater tolerance can be allowedbetween the starting and operating positions of the arc stream. The depression of the arc during starting, while larger than that'occurring with the structure of Fig. 2 is still extremely small as contrastedwith the structures heretofore employed, due to the steepness of the field gradient, with the observed re-' sult that the arc is never forced into contact with the lower wall of the arc tube 5 even with the simplified structure illustrated in Fig. 4.

In case it is desired to still further restrict the) deviation of the arc stream from the center of the arc tube l during the starting period the circuit arrangement shown in Fig. 8 is employed. In this arrangement the coils 6 am]. 8 instead of being connected directly in series with the arc, as in Fig. 6, are connected in series therewith through the series transformer M. The iron of this transformer is designed to be substantially saturated by the normal arc current, with the 119 result that the current flowing through the coils l5 and 8 increases very little during the starting period. Furthermore, during the continued fiow of the abnormal starting current through the transformer [Hi the current flowing through the coils 6 and t is somewhat out of phase with the arc current. As a result of both these factors the magnetic field in the arc tube 5 is maintained at a value very little in excess of that existing under equilibrium conditions, so that the forces tending to abnormally depress the arc during the starting period are greatly minimized with this novel arrangement. While illustrated in connection with the magnet structure shown in Fig. 2 it is obvious that this use of a saturable series transformer is of even more value where the magnet structures of Fig. 3or 4 are employed, the movement of the are below the tube axis being especially decreased in either of those cases. Where this transformer it is used the ratio may to be made whatever desired, it being preferable in some cases to use higher currents and fewer turns for the coils t and 8. Of course, where this transformer is employed it is possible to use less in.- sulation on these coils ti and 8, making it easier to confine them to the small area which gives the steepest field gradient.

In case it is desired to dispense with these cores and magnet coils I have found that the same results can be attained, although to a lesser degree, with the structure illustrated in Fig. 9. As shown in this figure therev is a heavy conductor i5 arranged longitudinally along the top of the lamp l and substantially in contact therewith, and a second conductor l6 similarly arranged below the lamp l. These conductors, which may be made J as vanes and polished so as to absorb the minimum amount of light, are connected in parallel across the secondary of the series transformer it which is connected in the arc circuit. The con.-

nections are such that the current flowing in, these conductors l5 and i6 is in the opposite direction to the arc current, as indicated by the arrows. The return wire i8 is arranged at some distance from the lamp i so that it will not affect 5 the arc to any appreciable extent, or it may be magnetically shielded. The transformer IT! is necessarily employed in order to give.the high current, of the order of ten to fifteen times the arc current, which is required by-each of the wires l5 and t6. In addition this transformer may have a saturable core, as described in con-- nection with Fig. 8, in order to decrease the abnormal arc depressing effects of the high current in the are at the start of operation. 65

This novel structure, ofcourse, provides a reversing field, with a point atv which there is no field whatever, in somewhat the same manner as the structure of Fig. 2. The field gradient is not particularly steep, however, and hence the arc varies through a considerable range inthe arc tube 0 between starting and operation, although it can never come into contact with the lower envelope wall due to the reversed field. The bowing of the arc is, of course, not minimized with this structure 'as it is with the structure of Fig. 2, but in some cases, this is not objectionable. In order for the arc to operate along the axis of the arc tube 5 under normal conditions it is essential that the conductor I6 should be somewhat smaller than the conductor. l5, so that the current will divide unequally therebetween, or that it should be somewhat spaced from the bottom of the lamp l, or both, so that there will be a field of the necessary intensity at the middle of the arc tube 5. In some cases, where it is not essential to have the arc control exertedby this reversing field, the lower conductor [6 may be omitted, in which case a smaller current in the conductor I5 will sufiice. In this case the are will wander farther below the axis of the tube during starting, but will not come in contact with thebottom wall of the tube, especially if the iron of the transformer I1 is saturated by the normal arc current. In the latter case it is even possible to bring back the return wire l8 along the bottom of the lamp l, despite its attraction for the arc, the current in this circuit then being about half that necessary with a single conductor along the top.

From the foregoing it will be apparent that important features of my invention are theuse of a magnetic field with an extremely steep, and preferably a reversing, gradient within the arc tube of a high intensity vapor arc lamp to confine the are at all times to a limited zone within said are tube, and to the use of means to limit this magnetic field during starting to even further restrict this zone within which the arc operates, and also to means to reduce or eliminate the bowing of the are.

While I have illustrated my invention by reference to certain specific structures which I have found to produce successful results it is obvious that my invention is not limited thereto, but that various changes, omissions and substitutions. within the scope of the appended claims, may be made therein or in the mode of operation without departing from the spirit oi' my invention. It is furthermore obvious that the various steps of my novel mode of operation may also be carried out with other types of apparatus, by hand, or by timing mechanism, especially where the gaseous discharge device is used on direct current, when separate fields, independent of the arc current, can be employed.

I claim as my invention:-

1. An electric gaseous discharge device comprising a sealed envelope containing a gaseous atmosphere and having electrodes sealed therein with the path therebetween at an angle to the vertical, in combination with means to produce a magnetic field within said envelope, along substantially the entire path between said electrodes. the gradient in said field being such that the magnetic reaction thereof with the normal are current is sufilcient to maintain the arc stream substantially at the axis of said device while that with the starting current is insufilcient to force said are stream into contact with the lower wall of said device.

2. In combination, an electric gaseous discharge device comprising a tubular sealed envelope containing a gaseous atmosphere and having electrodes sealed therein, the axis of said envelope being at an angle with the vertical, and means to produce a magnetic field within said device the intensity of which is markedly difl'erent on opposite sides of the axis of said envelope, the intensity of said field at the axis of said tube being such that the magnetic reaction thereof with the normal arc stream flowing along said axis is substantially equal to the forces tending to move said are stream away from said axis, while the intensity of said field drops I v said electrodes at an angle to the vertical, and

means to produce two opposing magnetic fields 15 within said device in a direction normal to the discharge path between said electrodes, the relative intensities of said fields being such that they neutralize each other at a point'slightly below the axis of said tube. I

4. In combination, an electric gaseous discharge device comprising a sealed envelope containing a gaseous atmosphere and having electrodes sealed therein, and means to produce a magnetic field within said device to control the position of the arc therein, said means comprising a pair of magnet coils above said device and connected in series with said are, one of said cells having a core with a small air gap therein which is located in proximity to said device, whereby the field therefrom has a steepgradient within said device. while the other coil is more remote and has a core with a large air gap, whereby the, field therefrom is more nearly constant throughout said device, said fields being opposed; and the relative intensity thereof such that the resultant field produced thereby is zero at a point slightly belowthe axis of said device.

5. In combination, an electric gaseous discharge device comprising a sealed envelope containing 40 a gaseous atmosphere and having electrodes sealed therein, and means to produce a magnetic field within said device to control the position of the arc therein, said means including a magnet coil and a core therefor, said core extending along said device beyond each oi said electrodes, 'the central portion of said core having a greater magreluctance than either end thereof.

6. In combination, an electric gaseous discharge device comprising a sealed envelope containing so a gaseous atmosphere and having electrodes sealed therein and means to produce a magnetic field within said device to control the position of the are therein, said means being connected in series with the are in said device through a series transformer, the core of said transformer being substantially saturated by the normal arc current, whereby said field is maintained substantially constant irrespective of increases in the are current. a

'7. In combination, an electric gaseous discharge device comprising a sealed envelope containing a gaseous atmosphere and having electrodes sealed therein, and a conductor extending along said device above the are path between said elec- 65 trodes, said conductor being connected in series with the are in said device with the current flowing therethrough in the opposite direction to that in said are.

8. In combination, an electric gaseous discharge device comprising a sealed envelope containing a gaseous atmosphere and having electrodes sealed therein the discharge path between said electrodes being at an angle with the vertical, and a pair of conductors extending along said device 7 enemas b'nejabove and the other below the arc path hetween said electrodes, said conductors being connected in series with the are in said device through a current transformer with the current flowing through each'thereof in the opposite direction to that in said arc, said transformer having a ratio such that the current in said conductors is many fold that in said are.

9. In combination, an electric gaseous discharge device comprising a tubularsealed envelope containing a gaseous atmosphere and having electrodes sealed therein, a reflector mounted with a portion thereof closely adjacent to said device,

and means to maintain the are within said device in a desired. relation with respect to said refiector, said means comprising a magnet in back of said reflector and having a core with an air gap thereinsaid air gap being located in juxta-' position to that portion of said reflector which is adjacent said discharge device.

CARL E. m.

Images