US2351914A - Luminescent tube system and apparatus - Google Patents

Description

jm@ 2(9) Q. P, @@UQHER @TAL LUMINESCENT TUBE SYSTEM AND APPARATUS- Filed July 14, 1.941

Patented June 20, 1944 LUMINESCENT TUBE SYSTEM AND APPARATUS Charles ick August Philippe Boucher, Paterson, and Freder- Kuhl, Ridgewood, N. J., assignors to Boucher Inventions, Ltd., Washington, D. C., a corporation oi' Delaware Application July 14, 1941, Serial No. 402,412

Claims.

Our invention relates to fluorescent tube lighting, and more particularly concerns a method and manner of operating the present-day hot cathode fluorescent gas discharge tubes, which are now widely available on the market, on cold cathode operation.

The advantages which are attendant upon the operation of fluorescent gas discharge tubes on cold-cathode operation, with energy provided by high leakage reactance transformers which automatically provide the voltages required for satisfactory operation, including both the required high starting voltage and the lower operating voltage necessary once the are is struck, have already been touched upon in our two (zo-pending applications, Serial No. 402,410, iiled July li, .1941, and Serial No. 402,411, filed July 14, 194i, and now Patent No. 2,317,844, granted April 27, 1943. Only casual mention is made there, however, as to the feasibility oi the satisfactory use in such cold systems of tubes which are readily avallabie on the market but which are initially designed for hot-cathode operation, and the consequent achievement of material saving in initial investment. No manner of connecting rior mode or operating the hot-cathode 'tubes is described in either said application.

An important object of our invention, therefore, is to provide for the satisfactory employment cf hot-cathode fluorescent gas discharge tubes on cold-cathode operation. Another object is to produce a socket for the reception of a hotcathode fluorescent gas discharge tube in such manner that the tube will perform satisfactorily on cold-cathode operation. A still further object is to produce an electrical energizing system for the satisfactory employment of uorescent gas'discharge tubes on cold-cathode operation, whereby the tubes have a long life, are. quick starting, will both start readily and operate steadily under cold weather conditions, will function satisfactorily with dimmer operation, and which system involves a lighting unit characterized by its simplicity, compactness, low fixture cost, and low assembly and operating costs, and in which system iilament-breakage where hotcathode tubes are employed is a matter of no importance, and which is characterized by its sturdiness and reliability under all operating conditions.

Other objects and advantages in part will be obvious and in part pointed out hereinafter.

Our invention accordingly resides in the several elements, features of construction, and operational steps, and in the combination of each of the same with one or more of the others, all as pointed out hereinafter, and the scope of the application of which is set forth in the appended claims.

In the drawing,

Figures 1 and 2 are front and side elevations, respectively, of a high leakage reactance autotransformer which can be employed in connection with our new system.

Figure 3 is an exploded perspective view showing one form of socket which may be employed according to our invention, as well as the manner ln which the same is associated with the terminals of a typical hot-cathode fluorescent gas discharge tube which can now be readily purchased on the market.

As conducive to a more thorough understanding of our invention, we may say at this time that the use of nuorescent gas discharge tube lighting has sprung into considerable prominence during the last few years in various industrial and commercial elds, as well as in household lighting. Such fluorescent tube lighting in fact shows evidence of substantially replacing the more conventional incandescent ilament lighting in many ields of illumination.

Many reasons may be marshaled for this somewhat phenomenonal success of uorescent lighting. Among the more important of these may be listed the high luminous elciency of fluorescent tubes, whereby they produce the same or even greater light output than do incandescent bulbs, for a materially decreased wattage input. Additonally, the fluorescent tubes are readily adapted to the direct emission oi light possessing desired spectral composition over a wide range of the visible spectrum. The tubes employed in such fluorescent lighting systems are found to have a useful life far in excess of that oi known incandescent bulbs, and this longevity, coupled with low operating costs, makes the use of such lighting highly attractive from a purely economic standpoint. Such tube lighting is found to be admirably adapted to either indoor or outdoor floodlighting, and is of particular value, for example, in afterdark illumination in factories and the like. Because of the much greater conversion of the energy input into visible light, and because of the greater light-emitting surface of such tubes, as compared to incandescent light bulbs, fluorescent tubes are found to operate at much cooler temperatures, and such heat as is emitted, consisting in but small part of radiant heat, can more readily be accommodated by persons in the neighborhood of such tubes. Addiimportant reasons can be the deserved popularity of ally decreased. The presence of the oxide-coated nlamentary cathode relied upon in the operation of such voltages becomes increasingly difficult, arc becomes unstable in Blackening of the tube in the trodes is found to occur, due to the depositing thereon of the coating material volatilized from the cathodes.

Although the preliminary heating of the filamentary and the its characteristics.

atively low potential gradient between the electrodes, with soon as they are energized, it requires at a minimum, some 6 or 7 seconds in whichto strike an arc across hot cathode tubes in fluorescent lighting systems. This represents an appreciable time lag, and in many installations and uses is The essential feature found t'o be a highly objectionable source of annoyance. Further, because such hot-cathode tubes as are currently in use are designed to operate on low voltage input, it is impossible to rcduce the impressed voltage to any material extent or to start or operate the tubes satisfactorily on what may be termed dimmer operation A further object of our invention, therefore, is to produce a system of operating hot--cathode tubes on cold-cathode operation in such manner that they have extremely satisfactory starting characteristics and are susceptible to satisfactory performance under dimmer operation, and in which system the occurrence of a broken filament in the fluorescent tube interposes no problem. A still further important object o1' our invention is to produce a fluorescent light system employing in satisfactory coldcathode operation, the known hot-cathode tubes readily availablev on the market, and in which system,

the tube are unimportant, terized by the small number of parts in the electrical unit comprising the tube energy transforming means, Wiring and reflector.

Referring now more specifically to Figures i and 2, wherein is disclosed a preferred embodimentof our invention, the system which we employ consists basically of a high leakage reactance transformer, a source of energy for said transformer, tubes operated by the transformer, and sockets for receiving said tubes.

The transformer may be either the autotransformer or regular type, i. e., in which the primary and secondary coils are electrically independent of each other. In the instance under discussion and purely for purposes of illustration, we elect to show and describe the construction and use of an autotransformer. Additionally, the transformer may optionally be of either the single or double secondary type. Again for illustrative purposes, we disclose and describe a transformer having double secondary circuits with a. tube operated in each secondary circuit.

of this portion of our invention is that the transformer be of the high leakage reactance type, so that no additional auxiliary will be required in the lamp unit referred to hereinbefore.

With this preliminary explanation, it will be seen that the transformer core is comprised of a longitudinally extending central leg IU, and outer legs II and I2 extending in parallel, -spaced relation to said central leg, one on each side thereof. These legs preferably have the same length and are closed on each other at their ends and adjacent their mid-points by end pieces I3, I3 and Il, I4 and mid-core portions MIMZ, respectively, extending from said outer legs to said central leg.

The mid-core portions MIMZ together with that portion of central leg disposed therebetween, divide the core 'into two groups of parallel metallic magnetic paths. In the first group of parallel magnetic paths a first path may be traced from the middle of central leg III through mid-core portion MI, to the left in Figure 1 along leg Il, down first end piece I3, to the right across central leg I0, back to MI.4 In the same group, a second parallel path is traced from M2 to the left along leg I2, up core portion I3, and across central leg I0, back to M2. Similarly, in the second group, a first magnetic path may be traced from central leg I0, up through core portion MI, to the right along leg II, down first end piece I4, across central leg I0. back to Ml. In the same group or second parallel magnetic path is traced from M2 to the right along leg I2, up through core portion Il, across leg I to M2.

In each group of lparallel magnetic paths, a pair of spaces is formed between the outer legs. end pieces, mid core portions and the central core portion. Each pair of spaces is disposed, one on each side of central leg I0.

Paired primary and secondary coils PI, SI and P2, S2, respectively, are disposed one pair in each group of parallel magnetic paths. The coils are disposed in said spaces, and are positioned around said central leg, with the primary coils on opposite sides of and adjacent the common leg.

Pairs of intermediate high leakage reactance magnetic shunts extend from outer legs II, I2 between each pair of primary and'secondary coils, toward but short of said central leg I0, forming air-gaps therebetween of high reluctance calibrated according to the particular load for which the transformer is designed. The purpose of these shunts will be described hereinafter. Thus shunts ShI and Sh2 extend between primary coils PI, SI, from outer legs II, I2, re-n spectively, towards but short of central leg III, providing therebetween respective air-gaps GI, G2. In like manner, shunts S713, Shll extend between primary and secondary coils P2, S2 from outer legs Il, I2, respectively, towards but short of central leg I0. Air-gaps G3, G4 are thus formed between shunts and central leg I0.

While as stated, our present invention includes the use of but a single primary coil, where desired, we here disclose the use of two primary coils. These primary coils may be connected across a source I5 of alternating-current electrical energy in either series or parallel connection, and in either adding relation, or opposed or bucking relationship. Where the primary coils are series-connected the primary voltage, assuming symmetrically-constructed coils, drops in equal steps through the two coils, so that the potential drop across each coil is but half of the total primary voltage. I'hus for a given number of turns of wire in the secondary coils, parallelconnected primary coils must contain twice as many turns as series-connected primaries, so that the volt per turn will be the same in either hook-up.

In the illustrative embodiment under discussion, we elect to connect the primary "coil in series-connection across the source of primary energy. Bucking relationship of the primary coils may result in slightly higher efciency inasmuch as when the electrical load across oneu vsecondary coil increases, there will occur with opposed relationship of the primary coils a shift in the primary ux from one parallel magnetic path to the other with attendant increase in the voltage induced in the other secondary coil. In the instance under discussion, we elect to connect the primary coils in opposed relationship.

With the following in mind, let us assume that the current for the given half-cycle under discussion is flowing to the right from source I5. Then a primary charging current may be traced through leads I6 and I1 to terminal I 8, thence to the right through primary coil P2, terminal I9, lead 20, to terminal 2I thence to the right through primary coil PI, terminal 22, and through leads 23, 24 back to the left side of the source I5. Of course, during the next alternate half-cycle of current ilow, the direction of the flow of the charging current through the primary circuit will be exactly opposite that just traced. The directions in which the primary coils PI and P2 are wound are opposed to each other, to give rise tothe bucking relationship referred to.

We have already suggested that it is within the compass of our invention to employ a transformer having but a single secondary coil, energizing but a single iiuorescent tube. Illustratively, however, we here employ a transformer having two secondary coils, each operating in separate circuit a fluorescent gas discharge tube. We vhave also suggested hereinbefore that we can employ our transformer in either regular or,

autotransformer connection. Again for illustrative purposes we have elected to illustrate and describe an autotransformer connection for voltages within the range of which the use of autotransformers is permitted by the ilre underwriters, and which range includes the operating potentials of practically all known hot-cathode fluorescent tube equipment. The use of autotransformer connections gives rise to certain advantages in low first cost and compactness of equipment.

Inasmuch as the primary coils are series-connected across source I5, and since the primary and secondary coils are autotransforrner-connected, we take advantage of this situation to increase the terminal voltage induced across the secondary circuits to values substantially the same as those which would maintain were the primary coils parallel-connected, by series-connecting each secondary coil with both primary coils.

Accordingly, each secondary coil is connected in series, ilrst with that primary coil positioned in the opposite magnetic path, and then with the primary coil in its own magnetic path. Thus, a secondary circuit may be traced from secondary coil SI through terminal 25, lead 26, through fluorescent gas discharge tube TI, and lead II to terminal I8 of primary coilPZ, thence to the right through said primary coil, through terminal I9 thereof to lead 20, thence to terminal 2|, to the right through primary coil PI, down through terminal 22, leads 23 and 24, to terminal 21 and back to the left through secondary coil SI, to the starting point. Similarly, a circuit may be traced from the secondary coil S2, from terminal 28 thereof down through lead 29 to fluorescent gas discharge tube T2, then through lead 23 and terminal 22 to primary coil PI in the opposite magnetic circuit, to the left through said primary coil, across terminal 2|, lead 20 and terminal I9 to primary coil P2 in the same magnetic circuit, then to the left through that primary coil, and through terminal I 8 and leads I1 and I6 to terminal 30 and thence to the right through secondary coil S2 to the starting point. It will be seen that each secondary coil contains a corresponding fluorescent gas discharge tube in circuit therewith and that additionally, each secondary coil is in series, first with the opposite primary coil and then with its adjacent primary coil. During the next subsequent half-cycle the direction of current ows through the two circuits of course would be the reverse of that just described.

The tubes TI. T2 are each of the hot-cathode type, having fllamentary electrodes led out to pairs of exteriorly extending terminals, which terminals are short-circuited for our desired coldcathode operation. The tubes are lined with suitable fluorescent salts, to give rise to the desired secondary visible radiation.

While it is not necessary, we prefer to design the two tubes of the same electrical characteristics, By such construction it is possible for the electrical constants of the several primary and secondary coils. respectively, to be the same, as is also true of the two parallel magnetic paths.

It will be interesting at this point to trace the passage of the primary magnetic ilux through the two parallel magnetic paths for an assumed half-cycle of charging current flow. Let us assume that the direction vof charging current is such that the primary flux courses in the direc.

tions of the arrows shown in Figure l. Then the primary flux from coil PI courses to the right along central leg I0. At the common magnetic leg comprised of mid core oprtions MI, M2 and central leg Ill, the primary flux from coil PI encounters and is lbuc'ked by flux from primary coil P2 tending to flow along central leg I0 in the opposite direction.

Choosing the paths of least reluctance, and splitting into two approximately equal streams because of the symmetry of the parts of the parallel magnetic paths under discussion, one stream passes up core portion MI, and courses to the left along leg II. Choosing the path of least reluctance, practically all of this primary flux avoids shunt Shl, and continuing to the left along leg II, courses down end piece I3 to leg I at the left of secondary coil SI.

At the same time the other stream of flux courses down core portion M2 and passes to the lett along leg I2. and up through end piece I3 to central leg I0. The two streams of ilux there re-uniting, the combined stream courses along central leg I0, across and interlinking secondary coil SI, back to primary coil PI. A high voltage is induced in secondary coil SI, which is under open-circuit conditions because the arc is not yet struck across tube TI, so that no back magnetomotive force is developed by this coil.

At the same time, primary flux from primary coil P2 courses the second group of parallel magnetic paths in the following manner: The rlux courses along leg I0 to the left of coil PI. In the mid core legs which are in common to the two groups of parallel magnetic paths the stream of flux bucks the primary i'lux from coil PI, and splitting into two streams which are substantially equal because of the symmetry of the paths comprising this magnetic path, one stream courses upward through core portion MI and to the right along leg I I. Choosing the path of least reluctance, by far the larger part of this stream of Ilux avoids shunt Sh3 and courses down end piece I4 to central leg I0. At the same time the second stream of flux courses down through core portion M2, to the right along leg I2, and up leg Il to central leg I II. The two streams of flux there re-uniting, the combined stream courses along leg III 'back to primary coil P2, and interlinking the secondary coil S2, induces therein a potential of high value. Since the arc across the tube in circuit with this coil is not yet struck, the coil has open circuit, so that no back magnetomotive force is developed therein.

During the next half-cycle of primary current fiow, the direction of coursing of the primary ilux is opposite of that described. For example, the ilux from primary coil PI courses to the left along leg I0, interlinking secondary coil SI and inducing a high secondary voltage therein. Splitting into two substantial equal streams, one

stream course upwardly opposite the arrows through end piece I3, to the right along leg II, down core portion MI, to leg Ill. The other stream courses down end piece I3, to the right along leg I2, and up core portion M2. The two streams re-uniting at central leg I0, the combined stream courses back to primary coil PI. Similarly, the primary ux from coil P2 courses the second group of magnetic paths to the right along leg Ill, interlinking secondary coil S2 and inducing high secondary voltage therein. The flux splitting into two substantially equal streams, one stream courses upwardly through end piece Il, to the left along leg II, anddown through core portion MI. The other stream continues down through end piece I4, to the left along leg I2, and up through core portion M2. 'I'he two streams of flux reunite at leg IU and course back to primary coil P2.

As stated at an earlier point herein, we prefer to design the transformer, where double secondary circuits are provided for, to operate with secondary loads of the same electrical characteristics. While this is preferable, from the standpoint of manufacturing simplicity and hence low first cost, it is not essential, and where it is desired to handle asymmetrical loads, this involves simply a question of proper design of the intermediate shunts, the selection of the proper wire sizes, etc., to achieve highest efficiency.

Even though the tubes TI, T2 are selected as supposedly having the same electrical characteristics, however, it almost invariably proves that the impedance of one tube-containing secondary circuit is slightly less than that of the other secondary circuit. Accordingly, after the passage of but a comparatively few current cycles, the high potential secondary current impressed across the terminals of the tube excites the gas molecules therein to a condition where they are ready to support an arc discharge. The arc immediately strikes, whereupon a back magne- Jomotive force is induced in the corresponding secondary coil. This induces a secondary flux which opposes the coursing of the primary flux.

Let us assume that it is the tube T2, which forming part of the secondary circuit of lower impedance, strikes rst. Secondary flux developed in secondary coil S2 opposes the interlinking of primary Ilux therethrough. The primary flux from coil P2 therefore seeks the paths of least reluctance. The design of the transformer core with respect to its associated windings, i. e. the proper design of the intermediate shunts, is such that as soon as the load across the secondary coil increases, only sulllcient primary Iiux continues to interlink the corresponding secondary coil as is necessary to induce therein a voltage suflicient to energize the increased secondary load. Inasmuch as the shunts Shi'. SM and their associated air-gaps G3, G4 now have a reluctance which is less than that of the path through secondary coil S2, some part of the primary flux from primary coil P2 is shunted through these elements and courses back to the primary coil.

When an arc strikes across tube T2, appreciable current iiow takes place, which courses not only through the secondary coil S2, but as well through the primary coils PI and P2 electrically connected thereto. This means that increased ilux is developed at such times by these primary coils, inasmuch as the iiux development is a function of the ampere-tunis. This greatly increased stream of primary ilux induces a secondary potential in coil Sl of such high value that the arc across tube TI strikes almost immediately. Promptly, then, a secondary flux is developed in coll SI, opposing the flow of the primary flux. A

Steady operating conditions now maintain, and the shunts ShI-SM and associated air-gaps GI--G4 are designed to permit just enough primary flux to interlink the secondary coils SI and S2 to induce therein the voltages required to maintain the arc discharge across tubes TI, T2. It is of course to be understood that less voltage is required to maintain the arc discharge than is needed to strike the arc.

The primary ilux may now be traced as iollows, assuming the charging current half-cycle to be such that the ilux courses generally in the direction of the arrows. From coil PI the flux .plits into two equal streams. One stream courses up along mid core portion MI, to the left along leg il, and a small measured proportion passes along, down first end piece I3 and back along leg l to primary coil PI, interlinking secondary coil SI as it goes. The major part, however, choosing the path of least reluctance, passes down through shunt Shi and across air-gap GI, back to primary coil Pi, completely lay-passing secondary coil SI. As to this part of the iiux the only energy losses are represented in the primary coil energizing current. The other stream courses down through mid core portion M2, to the left along leg I2, and a small quantity of the flux courses up second end piece I3 and back along central leg I0 to primary coil PI, interlinking secondary coil SI. It is these interlinking flux streams which induce in the secondary coil SI, the potential required to maintain the arc discharge across tube Tl. The greater part of the flux, however, courses up through shunt Sh2, across air-gaps G2, and back to the primary coil PI, completely by-passing secondary coil SI.

During the next alternate half-cycle of charging current, of course, the direction of coursing of flux is exactly the reverse of that described.

At that time flux will course to the left from primary coil PI, along leg I0. The required small quantity will continue along leg I0, interlinking secondary coil SI, and splitting, part will course up rst end piece I3, to the right along leg Il, down core portion MI, and back to primary coil PI. Simultaneously, the other stream will flow down second end piece I3, to the right across air-gap G3, back to central leg I0 and coil P2, completely by-passing coil S2. At the same time the second stream of iiux courses down through core portion M2 to the right along leg I2, and a small part continues up through second end piece I4, to central leg I0, and back to the left therealong to coil P2, interlinkng secondary coil'S2, during its passage. It is 'these two bodies of ilux which serve to energize secondary coil S2 during normal operation of the tube T2. The

larger part of this stream, however, courses up through shunt SM, across air-gap G4, and back to primary coil P2, completely shunting coil S2.

During the next alternate half-cycle of charging current flow, the paths of the stream of primary flux are the reverse of those traced, in a manner similar to that already described herein with reference to the first parallel magnetic path.

along leg I2, and up through core portion M2 to primary coil PI. The major part of the flux, however. before it reaches coil SI, splits into two equal streams, one of which courses up across air-gap GI, through shunt Shi, to the right along leg II, and down through mid core portion MI, back to primary coil PI. The other stream courses down across air-gap G2, through shunt Sh2, to the right along leg I2, and up mid core portion M2, to primary coil PI.

Similarly, during the first-mentioned charging half cycle, primary iiux courses to the left along leg I0 from primary coil P2. Bucking the primary flux from coil PI, in the leg common to the .two parallel magnetic paths, part of the flux courses upwardly through mid core portion MI, to the right along leg II, and a small quantity continues down through first end piece I4, and to the left along leg I0, back to the primary coil P2, interlinking secondary coil S2 during its'passage. The greater part of this stream of ilux, however, courses down through shunt S713, and

While we may employ a fluorescent gas discharge tube designed especially for cold-cathode operation and in which the incandescible filaments are omitted, and in which but a single terminal is led out of the tube at each electrode, these tubes are not available on the market, and can be obtained only at considerable expense, on special order. On the other hand the known hot-cathode type of fluorescent tube, having one or more llamentary electrodes, each with a pair of terminals projecting exteriorly of the tube, are readily available on the market, at comparatively low cost.

Accordingly, we have evolved an ingenious method of operating these known hot-cathode tubes on cold-cathode operation, which will now be elaborated upon, having reference to Figure 3. Here the tube consisis of a glass cylinder 3l' interiorly lined with a suitable fluorescent salt. Filamentary-electrodes, not shown, are disposed one at each end of the tube, and these electrodes are each connected at their ends to a pair of terminals 32, 33 which extend exteriorbr through a base 4I. This base, usually of metai, may be formed of suitable plastic, hard rubber, or the like.

The 'socket for receiving these terminals may follow any one of a variety of ypossible constructional embodiments. The essential point to be realized, however, is the short-circuiting of these filaments.

To illustrate, the socket may be somewhat of the bayonet type, in which the lamp terminals and base are inserted into the socket and are locked in place by a slight rotarytwist. The socket in such case is designed to short-circuit the terminals of the lamp. The'design may be such that at the same time, a connection of the primary coil across the source of energy is established. This last feature is developed in our co pending application Serial No. 402,413 led July 14, 1941, and will not be enlarged upon here. A special jumper may be designed for dropping into the base of the ordinary socket, and is held in place, short-circuiting the tube terminals, by the base of the tube when the latter is positioned in the socket.

For purposes of illustration, however, we disclose a socket indicated generally at 34 and comprising a shell 35 of metal or suitable plastic. Within this shell and disposed in insulated relation thereto are metallic conduits 36, 31 for receiving the prongs 32, 33, respectively, of the tube 3 I. A metal jumper 38 is disposed Within the shell, electrically connecting together the conduits 36, 31, and thereby short-circuitlng them. A lead 39 to the associated secondary coil extends from the unit consisting of the two conduits and the jumper to the exterior through a cap 40 on the shell 35. In this manner the unit consisting of the prongs 32, 33 and the filamentary electrodes are electrically connected together and form a single electrode. The base Il of the tube may consist either of metal or suitable plastic, just as the tube is found on the market. Since, however, this base is insulated from the prongs 32, 33, the question of whether it is formed of metal or of a plastic is a matter of but small moment. V Short-circuiting the terminals adapts the tube for cold-cathode operation, and permits obtaining the manifold advantages referred to hereinbefore.

It will be readily appreciated that the invention described herein gives rise to many practical advantages not hitherto possible of achievement with the equipment at hand. Our invention makes it possible to adapt tubes already available on the market for service in our new electrical system. with increased tube life, much quicker starting characteristics, and with highly stable arc discharge, even with prevailing cold weather.

The new system, employing hot-cathode tubes for cold-cathode operation according to the teachings of this application, is materially simplifed as contrasted with known tube lighting systems. Complicated and fragile auxiliaries such as starter switches, starter circuits, starting compensators, and simple ballasts are avoided, with attendant simplification of the lighting unit, decrease in assembly costs, and increase in sturdiness. We intentionally short-circuit the tube filament and take it out of participation as a lament in the operation of the tube. This failure of this filament has not effect on the operation of the tube itself.

Further to demonstrate the material simplification which may be achieved by the employment of our new system using hot-cathode tubes on cold-cathode operations is the fact that the transformer and any power-factor regulating condenser which we may elect to use can be mounted, as one single unit, directly upon the reflector which houses the tubes.

We find that, especially under steady operating conditions, the tubes employed in accordance with our present teachings, give a much larger saving in power demand in the tubes, for the same light output. For example, we find that under similar conditions, and with the same brilliance in light output, our system requires only 380 milliamperes, whereas the known hotcathode tube systems require 420 milliamperes.

We claim:

1. In a fluorescent gas discharge tube lighting system, the combination of a standard hotcathode fluorescent gas discharge tube having incandescible filamentary electrodes spaced apart from each other, and having pairs of terminals for each said electrode extending therefrom; and a cooperating socket receiving each pair of terminals and presenting shunt means thereacross, thereby adapting the tube for initial cold-cathode operation.

2. In a fluorescent gas discharge tube lighting system, the combination of a hot-cathode fluorescent gas discharge tube having incandescihle fllamentary electrodes spaced apart from each other, and having pairs of terminals for each said electrode extending therefrom; means for energizing said tube at a potential substantially exceeding the hot-cathode potential; and two sockets connected therewith receiving said pairs of terminals and containing means for short-circuiting the same thereby converting both the filament terminals and the filament cooperating therewith into a single terminal and adapting the tube for cold-cathode operation.

3. A fluorescent gas discharge tube lighting system, comprising two hot-cathode fluorescent gas discharge tubes, each having two spaced filamentary electrodes, and a pair of terminals for each electrode extending therefrom; and a high leakage reactance transformer having a primary coil and two secondary coils capable of impressing across the electrodes of said tubes a potential substantially in excess of that for hot-tube operation to give satisfactory cold-cathode operation; and two pairs of shunt sockets connected with the transformer secondary coils and receiving the electrode terminals of said tubes and short-circuiting the same, thus adapting the tubes for cold-cathode operation.

4. An electrical system comprising a high leakage reactance transformer having a primary and a secondary coil therein and shunt means for by-passing a portion of the primary flux about the secondary coil when the load on the latter increases; a source of alternating-current electrical supply connected across said primary coil; a hot-cathode fluorescent gas discharge tube having flamentary electrodes spaced apart from each other and a pair of terminals for each said electrode extending through said tube and exteriorly thereof; and shunt sockets connected with said secondary coils, each socket receiving a pair of filament terminals and short-circuiting them, to adapt the tube for cold-cathode operation.

5. A fluorescent gas discharge tube lighting system comprising: two fluorescent gas discharge tubes of the hot-cathode type each having two spaced iilamentary electrodes, and terminals extending therefrom corresponding in pairs to the electrodes; means impressing across the electrodes of said tubes a potential substantially exceeding the hot-cathode potential of the tubes and suilicient for instantaneous starting and cold-cathode operation; and means receiving and electrically short-circuiting the individual pairs of terminals externally of the tube, thus adapting the hot-cathode tubes for cold-cathode operation.

CHARLES PHILIPPE BOUCHER. FREDERICK AUGUST KUHL.

Images