US2835840A - Resonance lamps for very low voltages - Google Patents

Description

y 0, 1958 M. LAPQRTE A 2,835,840

RESONANCE LAMPS FOR VERY LOW VOLTAGES Filed Dec. 7, 1956 INVENTOR MARCEL LAPORTE lfwzm ATTORN E) 5 ats RESQNAllCE LAMPS FOR VERY LGW VOLTAGES ltlarcel Laporte, Paris, France, assignor to Centre National de la Recherche Scientifique, Paris, France, a corporation or France It is known to use, for lighting purposes, cylindrical tubes which are generally straight, containing mercury vapour and a rare gas under a pressure of a few millimeters of mercury, the gas being especially argon or a mixture of rare gases, for example argon-krypton.

An electrode is mounted in each extremity of the tube; these two electrodes are identical, each of them being connected to one of the extremities of a filament which carries a coating of a material having a high thermo-electronic emission. This emission is produced by the passage of a heating current through the filament; it is only produced with the object of facilitating the starting of the discharge, and it is stopped by the automatic interruption of the heating current as soon as the ignition of the tube has been obtained. From that instant, the tube operates in autonomous discharge, that is to say without any supply of external energy other than that which is supplied in the circuit of the discharge itself.

it is known that, in accordance with this. type of dis charge, the characteristic v:f(i) of the tube is negative, which means that an increase in the intensity of the current produces a reduction in the voltage of operation.

The light emitted by these tubes comprises on the one hand the radiations emitted by the mercury vapour in the visible spectrum, and on the other hand, the radiations emitted by the fluorescence of the substance deposited a thin layer on the internal wall of the tube. This fluorescence, with a continuous spectral base, is excited by the ultra-violet radiations emitted by the mercury vapour, and especially by the radiation of resonance of the mercury having a wave-length \=2.537 Angstrom units.

An economic production of light is only obtainable with these tubes if they have a minimum length of several decimetres which make it necessary to have a maintenance supply voltage greater than 100 volts. By reason or" the length of the tubes, the current supply leads must be separated in order to be joined to the electrodes; in order to conceal these leads, it has been the practice to place them inside a casing arranged parallel to the tube over its entire length.

The ignition of these tubes necessitates in addition an auxiliary device for pre-heating the filaments and for the creation of a momentary surge of voltage; finally, their supply circuit must necessarily include a stabilising impedance by reason of the fact that they have a negative characteristic. A power which is of the order of 20 to 25% of that which is consumed in the lamp is expended as a pure loss in the stabilisation device.

The lead-containing casing, the auxiliary ignition and stabilising devices (the purchase price of which is several times that of the tube) cannot be easily moved by reason of their weight and their bulk, and the result is that fluorescent tubes are generally fixed in permanence to walls or to ceilings.

The present invention provides a remedy for the various lrawbacks which have been referred to above; it re lates to lighting lamps which will be known as resonance lamps inthe text which follows, by reason of the fact that there is employed in them the propagation from place to place and in all directions, of the radiation of resonance of the mercury vapour; the ordinary commercial tubes, as at present known, will be termed fluorescent tubes.

In the drawings:

Fig. l is a schematic sectional view of the invention.

Fig. 2 is a circuit showing an alternate form of the invention.

Fig. 3 is a graph showing the relation between current and voltage.

The resonance lamps comprise a glass chamber, the internal wall of which is coated with a fluorescent powder; their shape is no longer necessarily tubular; it may be of considerable variety-bulbs, balloons, plates, etc., or all shapes which are compatible with a mechanical strength sufficient to withstand the force of atmospheric pressure. These chambers will contain mercury vapour and argon (or as the case may be, a mixture of argon and krypton), at a pressure which will be suitably chosen to give the optimum working of one type of lamp; this pressure will in general be lower than that employed in fluorescent tubes and may be for example in the vicinity of one millimeter of mercury only.

Inside the glass chamber, there will be arranged in proximity to each other, that is to say at a distance which will not exceed a few centimeters, a cathode having a high thermo-electronic emission, preferably indirectlyheated and of the type with a sintered cathode, and an anode formed preferably by one or a number of grids arranged so as to collect to the maximum extent the electrons emitted by the cathode. It is essential to observe that the cathode of a resonance lamp must be heated for the entire duration of working of the lamp and not only at the moment of ignition as is the case with ordinary fluorescent tubes.

In these conditions, as long as the potential established between the electrodes is less than 4.6 volts, which po tential corresponds to the minimum excitation of the mercury at the first level of resonance, no emission of light can be observed and only a very small thermo-electronic current is set up in the lamp. When the voltage begins to exceed 4.6 volts, at certain number of atoms of mercury are excited by electronic shocks at the level of resonance, and the return of these atoms to their normal state is accompanied by the emission of the radiation x=2.537 Angstrom units, which excites the luminous fluorescence of the powder.

As the potential difference between the electrodes is progressively increased, there is observed, at a certain critical voltage V a sharp increase in the intensity of the current passing through the lamp and a corresponding sharp increase in the light emitted, a phenomenon which gives the impression of an abrupt ignition of the lamp.

This increase in current is due, to a major extent, to the fact that at the voltage V which will be called the ignition voltage, the electronic shocks begin to cause the appearance of positive ions; these ions are attracted towards the cathode and neutralise the space charge due to the electrons; the thermo-electronic emission, which was previously very small, then increases sharply with a corresponding increase in the number of electronic shocks capable of exciting the mercury vapour.

Experience has shown that the voltage V, of ignition ot' the cold lamp is in the vicinity of 12 volts, which is slightly greater than the excitation potentials of argon at its resonance level or at the adjacent meta-stable levels. The explanation of this ignition is to be found in the fact that the excited atoms of argon can ionise atoms of mercury by shocks of a secondary nature; the positive ions of mercury formed by this means, reach the cathode and more or less completely neutralise the electronic space charge which tends to hinder the emission.

If, as a result of its previous operation, the lamp is still hot enough, the pressure of the mercury vapour may be sufliciently high for the probability of collision between the electrons and the mercury atoms not to be negligible; a recently extinguished lamp can be re-ignited at a volt age only very little greater than 10.4 volts, which corresponds to the potential of direct ionisation of the mercury vapour.

In the case of a lamp which is still hot and with a cathode having a very high electronic emission, it is even possible to obtain a' re-ignition at a voltage only very little greater than 5.4 volts, which is the voltage corresponding to the excitation of an atom of mercury at a meta-stable level; it is only necessary that such an excited atom should be subjected to a new collision with an energy of 5 electron-volts, to be ionised; if such cumulative collisions are sufficiently numerous, the space charge surrounding the cathode will be neutralised and the re-ignition of the lamp will thus be obtained at a voltage only a little greater than 5.4 volts.

It is important to note that if the space charge is completely neutralised, the thermo-electronic emission from the cathode cannot exceed a limiting value which is determined, in accordance with Richardsons law, by its surface area and its temperature. In addition, the cur rent which passes through the lamp is the sum of the thermo-electronic current and the ionisation current; the latter cannot have a high value since, at the low voltages employed, each electron in its passage between the cathode and the anode can only ionise a very small number of atoms; it is thus necessary to ensure that the current which passes through a resonance lamp cannot exceed a limiting value which is only a little greater than the saturation current which the cathode would supply, in a vacuum, at the same temperature. This provision has been confirmed experimentally in the curve of the characteristics, and by way of example, there has been shown in Fig. 3 one of the characteristics obtained by experiment.

The-necessity of use of an auxiliary rare gas, and the reason for a favourable choice of its pressure can be explained in the following way: when the resonance lamp is at the ambient temperature, the pressure of the mercury vapour in it is very small, and the free mean path of the electrons, in the mercury vapour, is greater than the distance from the cathode to the anode; under these conditions, in the'absence of rare gas, the probability that the electrons emitted by the cathode can ionise the mercury vapour is too small for ignition, which necessitates the arrival on the cathode of a sufiicient number of positive ions, to take place. If the argon is introduced into the lamp at a pressure of the order of one millimeter, much greater than that of the mercury Vapour, the probability of excitation of the argon will no longer be negligible, as the ionisation of the mercury may be effected by collisions of a secondary nature with the excited argon atoms, the free path of which in the mercury vapour is much smaller than that of the electrons and which, in addition, are not controlled by the electric field.

Experience has shown that in order to obtain operation of a resonance lamp with a voltage in the vicinity of 12 volts, there exists an optimum value of the pressure for every anode-cathode distance; for example, for an anodecathode distance of 15 millimeters, the optimum pressure is in the vicinity of one millimeter of mercury.

The necessity for limitation of the pressure of the argon would appear to result from the losses of energy to which the electrons are subjected in their elastic collisions with the atoms of argon; if the pressure is too high and if, accordingly, these collisions are too numerous in the path from the cathode to the anode, the electric fieid will not be able to give them the energy necessary for the direct or indirect ionisation of the mercury vapour.

Experience has also shown that the characteristic of a resonance lamp is very different from that of ordinary fluorescent tubes. This result is not surprising since their methods of operation are totally diiierent: fluorescent tubes operate with an autonomous discharge, whilst resonance lamps operate with a semi-autonomous discharge; the current which passes through them falls in fact to zero immediately under the low voltages employed, if the heating of the cathode is stopped, the energy of this heating not being supplied in the discharge circuit.

The form of the characteristic v=f(i) of a resonance lamp depends on the geometric parameters which define the relative arrangement of the electrodes and, for any given lamp, on a number of other factors:

(1) On the intensity of the thermo-electronic current which the cathode is capable of supplying, that is to say on the nature of the said cathode, on its surface area and on the temperature to which it is brought;

(2) On the Working temperature of the lamp which determines the pressure of the mercury vapour; this pressure is a function of the power expended in the lamp and on the conditions under which it is cooled;

(3) On the pressure of the rare gas.

Tests in connection with the influence of these various factors have shown that the characteristics which are first of all negative in the case of low currents become clearly positive at currents above a certain value.

The very important result of this study has been to establish that it is possible to define, by the choice of the pressure of the argon and by the choice of the heating voltage for the cathode, a field of current intensities in which the characteristic is very clearly positive and in which, in consequence, stable operation can be obtained either without a stabilising member or, if too great variations of the supply voltage are to be feared, with a stabilising member of very small bulk which consumes only a very small fraction of the total energy of the supply. For example, for a type of lamp having a sintered cathode, the heating of which absorbs 13 Watts, in which the cathode-anode distance is 15 millimeters and the pressure of the argon is 1.26 mm. of mercury, the current intensity only varies from 2.2 to 2.6 amperes when the voltage between the electrodes varies from 8.7 to 104 volts: this lamp being ignited from the cold state at about 10 volts, it will be seen that its operation does not require any stabilising member if a generator is employed, the electro-motive force of which is only slightly greater than 10 volts.

In a resonance lamp, the emission of the radiation of resonance is first of all produced in the space comprised between the cathode and the anode, but it is known that these resonance radiations can be absorbed by adjacent atoms and then re-emitted by them in any directions. The radiation of resonance is thus diffused from place to place, and this in all directions, and it can thus reach the walls of the chamber and illuminate the fluorescent deposit on the walls, whatever the form of these walls may be. Objects of any shape, placed inside the lamp and covered with a suitable fluorescent deposit can also be illuminated; they can be seen through parts of the walls of the lamp which are not themselves provided with a fluorescent deposit. It is clear that this possibility can be used for publicity purposes and for new decorative elfects.

The anode and the cathode of a resonance lamp being arranged at a short distance from each other, the input and output passages for the current may be effected in the same base; on the other hand, by suitably chosing its resistance, the heating filament may be connected in parallel between the cathode and the anode; the complete supply of the lamp can then be efected by means of two wires very close together: resonance lamps will thus have the advantage over tubular lamps, which up to now was reserved for incandescent lamps, of constituting portable sources of light whilst at the same time they retain the luminous qualities peculiar to fluorescent sources.

By way of indication and without any sense of limitation, there will now be described a number of forms of embodiment of the invention, reference being made to the diagrammatic drawings shown in the accompanying Figs. 1 and 2 and which concern respectively a lamp 'ng the shape of a bulb as currently employed, and an. alternative form of this device which comprises two anodes and two cathodes, with the object of reducing the flickering of the flow of light when the lamp is supplied with alternating current.

Referring now to Fig. l, the device comprises a glass bulb 1 of any desired shape, coated internally with a fluorescent substance 2 and provided with a pipe 3 which is necessary for filling the bulb with rare gas and mercury vapour, after having produced a very high vacuum in the bulb. The device also comprises an oxide cathode 4, preferably of the sintered type with indirect heating, provided by means of a heating filament 5. The anode 6 is preferably in the form of a grid.

These members are connected to the conducting wires 7 and 8 as shown in Fig. l, and finally these conductors are connected to the supply circuit, which may comprise an inductance coil or regulating resistance 9 which may be dispensed with if too great variations in the supply voltage not to be feared. The supply circuit also comprises a low-tension source of alternating or directcurrent voltage of about 12 volts, applied across the terminals Hi and ii.

When the lamp is supplied at a constant voltage, for example by means of an accumulator battery, the light radiation is constant. In the case in which the lamp is supplied from an alternating voltage source, for example at a frequency of 50 cycles per second, there is observed in the case of a system of connections as shown in Fig. l, a slight flickering which results from the fact that the charge only passes during a fraction of one of the halfwaves of voltage.

This defect is very small and is eliminated in practice if the lighting is obtained by means of a double circuit employing two lamps, the anode of one being connected to the pole of the secondary of the transformer to which is connected the cathode of the other lamp. The same result can be obtained with a single bulb con-- taining two anodes and two cathodes, for example in accordance with the circuit shown diagrammatically in Fig. 2.

in this figure, the two filaments l2. and 13 are connected in parallel; during one half-wave, the discharge passes from the anode 14 to the cathode 15 which is heated by the filament 12, and in the next half-wave, from the anode 16 to the cathode 17, which is heated by the filament 13; the lamp then operates at a frequency which is twice that of the supply system.

It will be observed that in this diagram the electrodes 14 and 15 are placed far apart from each other; the same condition holds for the electrodes 16 and 17, the anodes M and 16 respectively protecting the cathodes 15 and 1'7. In addition, the ignition is controlled by the switch 13 connected in the circuit of the secondary of the trans former 19, which is not obviously in any way combined with the lamp.

In all these devices, the anode may comprise a number of grids suitably spaced apart in order to ensure an almost complete interception of the electrons.

are

It should be noted that the ignition of the discharge of a resonance lamp is only produced when, on the one hand, the cathode is brought up to a temperature which makes its electronic emission sufficiently intense and when, on the other hand, the pressure of the mercury vapour has reached a sufiicient value by reason of the heating-up of the chamber. For this reason, a certain time elapses between the instant at which the current is applied and the instant corresponding to the production of the normal lighting at full intensity, the duration of this delay being reduced directly with the heat capacity of the cathode. It has proved a simple matter to choose this capacity in such manner that the delay does not exceed a few seconds. In addition, during this time, and almost instantaneously with the closure of the current circuit, there is produced a progressively increasing lighting effect due to the emission of light of thermal origin from the heating filament of the cathode, and then due to that of the body of the cathode, as their temperature rises. There is thus obtained in a very agreeable mannor, a switching on with progressive lighting effect which starts almost at the same time as the switch is closed and which reaches its full effectiveness in a few seconds. The reverse phenomenon of a progressively decreasing light is obtained when the lamp is extinguished.

By reason of the fact that the cathode is heated to a fairly high temperature in a gaseous atmosphere at low pressure, it is preferable to prevent as far as possible the formation of an opaque deposit on the walls of the lamp, due to vaporisation of the metallic support of the cathode. T 0 this end, the non-emitting parts of the cathode may be enclosed by a sleeve, for example of quartz, which will have in addition the advantage of reducing the power for heating.

What I claim is:

1. In a resonance lamp for producing fluorescent illumination and adapted to operate from both alternating current and direct current sources at potentials less than 15 volts, the combination including an external envelope at least a portion of which is transparent, fluorescent material on the internal wall of said envelope, said material having a continuous spectral base, mercury vapour and at least one of the rare gases within said envelope, the pressure thereof being about one millimeter of mercury, at least one thermionic emission sintered cathode in said envelope, indirect heating means for said cathode operable during the entire operating period of said lamp, and at least one grid-like anode in said envelope located adjacent said cathode at a distance of not more than a few centimeters therefrom, said fluorescent material being excited by the ultra-violet radiation and particularly the radiation of resonance of mercury having a wave length of 2.537 Angstrom units, the ignition voltage of the cold lamp being about 12 volts.

2. A resonance lamp as in claim 1 including a stabilizing impedance to compensate for potential variations in the supply to the lamp, and a base housing said impedance and leads to said cathode and anode.

References Cited in the file of this patent UNITED STATES PATENTS 2,030,805 Wiegand Feb. 11, 1936 2,182,732 Meyer et al. Dec. 5, 1939 2,222,668 Knoll Nov. 26, 1940 2,409,771 Lowry et a1. Oct. 22, 1946

Images