EP0543822B1

EP0543822B1 - Discharge lamp arrangement

Info

Publication number: EP0543822B1
Application number: EP91910739A
Authority: EP
Inventors: Antony Tambini; Ian Hopkin
Original assignee: GE Lighting Ltd
Current assignee: GE Lighting Ltd
Priority date: 1990-06-19
Filing date: 1991-06-14
Publication date: 1997-04-16
Anticipated expiration: 2011-06-14
Also published as: ATE151916T1; WO1991020092A1; DE69125717D1; DE69125717T2; GB9013677D0; EP0543822A1

Abstract

A discharge lamp assembly in which electromagnetic radiation is produced by a gas discharge in an envelope containing a fill. The charge carriers for initiating and sustaining the main discharge are supplied from a plasma created by a pilot discharge in the fill.

Description

The present invention relates to a low pressure discharge lamp, such as a fluorescent lamp, in which electromagnetic radiation is produced by a discharge in an envelope containing a fill.
A discharge in a discharge lamp conducts electricity because atoms in the fill have been ionized to produce electrons and positive ions. Ionization in a low pressure discharge lamp is principally caused by electron impact.
When a potential difference is applied across the discharge, the electrons are provided with energy and accelerated. In a conventional discharge lamp, the energy provided is sufficient to enable the electrons to ionize at least one constituent of the fill, producing more electrons which are in turn accelerated to cause further ionization and liberation of electrons. In addition, the ions are attracted to the cathode which thereby produces still more electrons by secondary emission and/or heating.
Thus, the electron density, and so the electrical conductivity, of the discharge increases with the current passing through the discharge. The increase in conductivity with current is normally so great that the voltage required to maintain the current in the discharge falls as the current rises i.e. the discharge has a negative voltage-current characteristic. Consequently, most discharges are not current limiting and, for stable operation from a constant voltage supply, must include a current-limiting device (a ballast), such as a resistor or, for a.c. operation, an inductor, a capacitor or some combination thereof which minimizes power loss and prevents current runaway.
Others in the lighting art have already sought to provide a discharge lamp in which the need for a ballast is at least reduced so that the size, weight and cost of the light source system as a whole (including control gear) may be reduced.
For example, EP-A-54959 (GTE Laboratories) discloses a fluorescent lamp in which the anode and cathode are separated by a distance which is less than the mean free path of the electrons in the fill, and in which the anode has an open mesh structure through which the electrons pass to form an electron beam. In this arrangement, the current drawn by the lamp is limited firstly by the very low probability of collisions in the region between the anode and the cathode, and secondly by the relatively weak electric field in the drift region beyond the anode which precludes ionization. For these reasons, although such beam mode lamps can be operated without ballast, their light output is very low.
JP62-12059A (Matsushita) discloses a lamp in which a first anode is again separated from a first cathode by a distance less than the mean free path of the electrons in the fill, and in which the anode has a mesh structure. In this arrangement, however, the electrons which pass through the anode ionize mercury atoms in the space beyond the anode, and this ionization neutralises the space charge around the anode. In these circumstances, electrons emitted by a second cathode in the space beyond the mesh anode are accelerated across the space toward a second anode to produce a luminous discharge. Advantageously, the voltage required to produce this discharge can be maintained below the ionization level, the mercury atoms simply being excited by the electron collisions. Again, therefore, the lamp can be operated without a ballast, but, as with the beam mode lamp, the light output is low.
In the present specification, there is described a fluorescent lamp that can be operated with reduced or zero ballast compared with existing lamps but which also has a comparable light output. In its compact form, the lamp is capable of replacing the conventional household incandescent lamp.
According to the present invention there is provided a discharge lamp assembly comprising: an envelope containing a fill capable of sustaining a discharge when suitably energised; a cathode and a first anode within the envelope; means for creating a pilot discharge having a positive voltage-current characteristic between the cathode and the first anode, the first anode being spaced from the cathode by less than the mean free path of electrons in the pilot discharge, the first anode allowing electrons of the pilot discharge to pass therethrough; a second anode spaced from the first anode by more than the mean free path of electrons in the discharge; and means for establishing a potential between the first and second anodes which accelerates the electrons which pass through the first anode towards the second anode to ionise the fill between the first and second anodes and promotes further ionisation of the fill whereby a main discharge is initiated and sustained between the first and second anodes.
In a preferred embodiment of the invention, the electrons produced by the pilot discharge in one region of the lamp are introduced into a second region of the lamp through a mesh electrode forming the first anode.
In operation, this second region contains the main discharge which exhibits characteristics similar to the discharge in a conventional discharge lamp in that:

(a) there is a high probability of electrons colliding with other particles in the fill as the second region has a dimension greater than the mean free path for electrons in the fill, and
(b) atoms in the fill can be ionised by collisions with electrons accelerated in the region, a sufficient proportion of the electrons being provided with an energy greater than the ionization energy of at least one constituent in the fill.

In contrast to the prior art, the inventors have discovered that the current in a main discharge producing electromagnetic radiation can be controlled by a pilot discharge having a positive voltage-current characteristic. In particular, the inventors have found that the full ionization voltage can be applied across the main discharge without ballast and without causing current runaway.
In essence, the prior art prevents current runaway by controlling the conditions of the discharge which produces electromagnetic radiation whereas the present invention effectively uses the pilot discharge as a controllable current-limiting cathode.
Preferably, the pilot discharge is created in a region having a dimension greater than the mean free path for electrons in the fill, and the pilot discharge is created by applying a voltage across this region such that electrons in the region are provided with sufficient energy to ionize atoms of at least one constituent in the fill.
Advantageously, the voltage for accelerating the electrons is applied between the cathode and the mesh electrode, the two electrodes being separated by a distance greater than the mean free path for electrons in the fill.
The mesh electrode or first anode is preferably interposed between the second anode and the cathode. In this case, the main discharge is struck between the mesh electrode and the cathode. The mesh electrode may have a cylindrical, planar or other form.
The cathode may advantageously comprise what is termed in the art as "a hot cathode". In this case the pilot discharge includes a ballast but it will be appreciated that the ballast requirements for a discharge lamp of a required light output will be much less than for a conventional discharge lamp because the ballast is required only to limit the current to a low power pilot discharge which in turn controls the current in the main discharge.
Alternatively, the cathode may comprise what is termed in the art as a "hollow cathode" with a positive voltage-current characteristic.
In contrast to the prior art, the positive voltage-current characteristic of the pilot discharge with either a ballasted hot cathode or a non-ballasted hollow cathode is not dependent on the prevention of secondary ionizing collisions in the fill. Accordingly, this removes the need for conditions which seem essential in the prior art beam mode lamps, such as pressures of the order of 0.1 mbar and close spacing of the electron emitting cathode and anode, but which are detrimental to the output of electromagnetic radiation from the lamp.
Embodiments of the present invention will now be described by way of example only, and with reference to the accompanying drawings in which:

Figure 1 shows schematically a first discharge lamp embodying the present invention;
Figure 2 shows schematically a circuit for use with the discharge lamp of Figure 1;
Figure 3 is a sketch of a floating potential map (measured with respect to the cathode) for the lamp of Figure 1;
Figure 4 is a graph showing electrical characteristics determined experimentally for the lamp of Figure 1;
Figure 5 shows a modification of the discharge lamp of Fig. 1;
Figure 6 shows schematically a second discharge lamp embodying the present invention;
Figure 7 shows schematically an operating circuit for use with the discharge lamp of Figure 6; and
Figure 8 is a graph showing electrical characteristics determined experimentally for the lamp of Fig. 6

Referring first to Figures 1 and 2, there is shown a fluorescent lamp 85 consisting of a glass tube 88 with a "hot" cathode 74 at one end and an anode 50 at the opposite end. The glass tube contains a fill of between 0.5 mbar (50N/m²) and 5.0 mbar (500N/m²) rare buffer gas, such as argon, and around 8 µbar (0.8N/m²) mercury. The cathode 74 is completely shielded from the anode 50 by a mesh electrode 54 which extends over the cross-section of the tube 88 and includes a flexible skirt 89 to ensure a complete seal. The mesh is mounted on an electrically conductive annular support ring 90.
A typical tube radius would lie in the range 10-40mm with a cathode-mesh separation l₁ of 2mm-100mm and a mesh-anode spacing of any desired length l₂ generally exceeding l₁. The mesh 54 preferably has a hole size in the range 0.05mm-1.0mm giving an open area of 30-50%.
The "hot" cathode 74 may be of the type normally used in fluorescent lamps, coated with an electron emissive material, and may be directly or indirectly heated. The tube 88 is illustrated as straight with a uniform cross-section but it can be bent into any desired shape, with the cross-section varying along the length if required.
The lead-in wires 91, 92 to each end electrode are sealed in the respective ends of the tube 88, and the lead-in wire 93 to the intermediate mesh electrode 54 may comprise an uninsulated wire within the tube.
In operation, a DC voltage Vapp is applied to the pilot discharge circuit consisting of the cathode 74, mesh 54 and a small ballast resistor R (Fig. 2). This produces a pilot discharge with a measured potential V₁ between the cathode and the mesh, and a pilot discharge current I1. The pilot discharge acts as a controllable current-limiting cathode for the main discharge between the mesh 54 and the anode 50. The inner surface of the tube 88 is coatd with a phosphor (not shown) to convert the ultraviolet radiation emitted by the pilot discharge and the main discharge to a luminous output.
A variable DC voltage V₂ is applied between the mesh and the anode, and I₂ is the (unballasted) main discharge current. Current I₃ is defined as the difference between the ballasted and unballasted currents, i.e. I₃ = I₁ - I₂. I₃ is therefore negative when the unballasted current I₂ is greater than the ballasted current I₁.
The pilot discharge appears to behave as a conventional ballasted hot cathode fluorescent discharge. Some electrons from this discharge pass through the mesh 54 and are influenced by the mesh-anode voltage V₂. These electrons are accelerated towards the anode 50, causing excitation and ionization to form the main discharge. This ionization results in a large ion flux back towards the mesh 54 and an ion sheath is built up on the anode side of the mesh. As clearly shown in Fig. 3, a large increase in potential known as the mesh fall is developed across this ion sheath; this is the principle control mechanism for the main discharge.
An increase in the main discharge current I₂ results in an increase in the mesh fall voltage so that the main discharge has a positive current-voltage characteristic as shown in Fig. 4. It is not possible to strike the main discharge if the pilot discharge is not already running.
Provided the mesh-anode voltage V₂ remains below the threshold value for arcing between the anode and the mesh, the main discharge remains stable over a wide range of voltage V₂ from the lowest that will sustain a discharge to values that will sustain a current at least comparable to that in a standard fluorescent lamp. In particular, the conditions within the pilot discharge (i.e. current I₁, voltage V₁ and the floating potential profile) are substantially independent of the main discharge voltage V₂. This is illustrated in Fig. 3 where the floating potential is plotted for two values of V₂ shown as V_2a and V_2b with V_2b > V_2a.
The intensity of the light output is generally dependent on the main discharge current I₂. Due to the mesh fall voltage preventing current runaway and providing a positive current-voltage characteristic as shown in Fig. 4, the current I₂ can be varied simply by varying V₂. Moreover, V₂ can be varied over a wide range without disturbing the stability of the main discharge. Accordingly, in contrast to conventional fluorescent lamps, the lamp can be progressively dimmed and brightened by varying the voltage V₂.
Alternatively, the current I₂ can also be increased by increasing I₁ so that dimming and brightening of the lamp could be controlled by varying V₁, for example by varying the resistance R.
Instead of extending the mesh 54 over the cross-section of the tube 88, the cathode 74 can be enclosed within a mesh cylinder. One such modification is shown in Fig. 5. In this modification, where like parts are denoted by the same references, the mesh 54 is replaced by a double mesh cylinder 54a, 54b, and the opposing ends of the cylinder are closed by a pair of insulating discs 96a, 96b. In this case, the inner mesh 54b may be of wide gauge aluminium with a hole size of 1.5mm to provide a backing support, and the outer mesh 54a of fine stainless steel, hole size 0.1mm. A double mesh could also be substituted for the single mesh 54 in Fig. 1, with the coarser mesh supporting the fine mesh.
The discharge lamp 85 therefore has a ballast-free main discharge controlled by a pilot discharge having a relatively small ballast. Efficacies produced at present, while less than those of conventional fluorescent lamps, are comparable with or greater than those of domestic incandescent lamps and, it is believed, can be increased by optimisation of lamp parameters.
Early prototype lamps 85 have typical efficacies of around 45 lm/W (this refers to total system Watts) with approximately 80% of the total power being unballasted.
An important consideration when operating such lamps is the power balance between the ballasted pilot discharge and the unballasted main discharge. Favourable operation generally requires that the unballasted current I₂ should be larger than the ballasted current I₁ (so that I₃ is negative). However, the conditions which produce a large negative I₃ also tend to result in a large mesh fall voltage; this is not always desirable since it represents a large internal power loss. A large mesh fall voltage creates more energetic ion impact to the mesh which leads to sputtering and a shortened life. Thus, the optimum conditions for operation can involve nearly equal current I₁ and I₂ with a low mesh fall voltage, coupled with a large value of V₂ so that the power dissipated in the main discharge itself is maximised. Particularly good results have been achieved with 1W dissipated in the ballasted discharge, compared to 12W in the unballasted discharge.
While the theory of operation is not yet firmly established, it is believed that the operation of the lamp can be summarised in the following way. A low power conventional ballasted discharge is formed in the pilot region 97 between the hot cathode 74 and the mesh 54. Due to sheath formation which covers the mesh holes, electrons approaching the mesh are subjected to a negative potential gradient, which acts as a selective filter in allowing only the higher energy electrons to pass through the mesh. The consequent reduction in electron current is then compensated by a large positive potential, referred to as the mesh fall, in a sheath on the opposite side of the mesh. Electrons entering the second discharge then have the full energy of the mesh fall, exciting and ionizing both mercury and noble gas atoms, to form the main discharge. The increased ionization in the negative glow region 98 of the second discharge causes a large ion flux to return to the mesh 54, which, due to the short collision length of the ions, is largely scattered on to and absorbed by the mesh. The combined effect of these processes is to allow the mesh to act as a control for the main discharge. Any increase in anode-mesh voltage V₂ is compensated directly by a similar increase in the mesh fall voltage and also causes the main discharge current to increase. Thus the main discharge has a strongly positive current-voltage characteristic, although the positive column between the mesh 54 and the anode 50 still maintains a weakly negative characteristic, which, as in the case of standard fluorescent lamps, produces most of the light.
Referring next to the "hollow cathode" embodiment of Figs. 6 and 7, the hollow cathode system 32 comprises a coiled tungsten hollow cathode 34 of approximately 4mm inner diameter concentrically surrounded by a stainless steel mesh cylinder 36 of inner diameter 15mm and hole size 0.3mm. The cathode 34 is supported by two ceramic discs 38a, 38b which also close the ends of the mesh cylinder 36. Lead-in wires 40, 42 connect the mesh cylinder 36 and the hollow cathode 34 to an electrical power supply (not shown); the lead-in wire 42 to the hollow cathode 34 is insulated to prevent cold cathode discharges in this region. The hollow cathode system 32 is positioned in a glass envelope 44 in a portion 46 of inner diameter 40mm and length 65mm. The portion 46 narrows to a discharge portion 48 of inner diameter 13mm and length 350mm, the internal surface of which is covered with phosphor (not shown). An anode 50 is positioned at the other end of the discharge portion 48. The glass envelope contains a fill of 2.5 torr (333N/m²) argon and 50 mg mercury. (The mean free path for electrons in such a fill is about 0.2mm).
An operating circuit for the lamp arrangement of Figure 6 is shown schematically in Figure 7. An auxiliary power supply 52 is used to apply a potential difference V1 between the mesh 36 and the hollow cathode 34 with the mesh 36 effective as an anode to the hollow cathode 34. The potential difference V1 is sufficient to generate a pilot discharge in a pilot region 54, the periphery of which is defined by the mesh 36. A potential difference V2 is applied between the mesh 36 and the anode 50 by a main power supply 56. A main discharge is produced in the region 58 of the envelope between the mesh 36 and the anode 50.
In both the pilot discharge and the main discharge collisions take place between electrons and fill atoms which result in ionization and excitation of the fill atoms. The mercury atoms emit electromagnetic radiation which excites the phosphor on the glass envelope to emit light. Both the pilot discharge and the main discharge were observed to emit electromagnetic radiation and to be very stable.
In examples of lamps 30 tested, the operating voltages were VI = 220V, V2 = 150V with currents being of the order of I1 = 70mA, I2 = 120mA (13=-50mA). A typical efficacy obtained was 27 lm/W.
The advantage of the hollow cathode discharge is that the cathode no longer requires heating, and it requires no ballasting. The theory of operation is described, for example, in "Studies of Metal Vapour Lasers" I.D. Hopkins PhD Thesis, University of Wales 1988 (and references therein).
The inventors have found that a variety of hollow cathode structures can be used to sustain the pilot discharge, including a helical coil and a tubular mesh.
Once a pilot discharge has been established in the hollow cathode 32, the electrons penetrating the mesh 36 are accelerated by the voltage V2 and cause a main discharge by ionization and excitation in the discharge region 58 between the mesh 36 and the anode 50. During normal operation of the lamp, provided V2 does not exceed the sparking potential (or glow maintenance potential in the absence of the hollow cathode 32) for the separation of the mesh 36 and the anode 50, the current I2 in the main discharge is controlled by the electron density in the pilot discharge and therefore by the current I1 for a very wide range of V2. The graph of Fig. 8 shows the relationship between the currents I1 and I2.
Modification to the embodiments described within the scope of the present claims will be apparent to those skilled in the art.
In particular, the illustrated discharge lamp arrangements may be modified for A.C. operation by having an electrode and mesh assembly at each end of a glass envelope.
For the avoidance of doubt, in the present specification, the term 'electrode' is used to indicate a structure which emits electrons, receives electrons or otherwise modifies the path of electrons or other charge carriers when a potential is applied thereto.

Claims

A discharge lamp assembly comprising:
an envelope (88) containing a fill capable of sustaining a discharge when suitably energised;

a cathode (74) and a first anode (54) within the envelope (88);

means (Vapp,R) for creating a pilot discharge having a positive voltage-current characteristic between the cathode and the first anode, the first anode (54) being spaced (l₁) from the cathode (74) by less than the mean free path of electrons in the pilot discharge, the first anode (54) allowing electrons of the pilot discharge to pass therethrough;

a second anode (50) spaced (l₂) from the first anode (54) by more than the mean free path of electrons in the discharge; and

means (V₂) for establishing a potential (V₂) between the first (54) and second (50) anodes which accelerates the electrons which pass through the first anode (54) towards the second anode (50) to ionise the fill between the first and second anodes and promotes further ionisation of the fill whereby a main discharge is initiated and sustained between the first and second anodes.
An assembly according to claim 1 in which the fill comprises mercury vapour and a rare gas.
An assembly according to claim 1 or claim 2 in which the inside surface of the envelope (88) is coated with a phosphor.
An assembly according to claim 1, 2 or 3 in which the envelope (88) is of elongate tubular form with the cathode (74) at one end and the second anode (50) at the other end.
An assembly according to claim 1, 2, 3 or 4 in which the cathode (74) is a hot cathode and the pilot discharge is ballasted to give a positive V-I characteristic.
An assembly according to claim 1, 2, 3 or 4 in which the cathode (74) is a hollow cathode to give the pilot discharge a positive V-I characteristic.
An assembly according to claim 4 or 5 in which the first anode (54) comprises a mesh (54) extending over a cross-section of the envelope (88).
An assembly according to claim 4, 5 or 6 in which the first anode (54) comprises a cylindrical mesh bounding the cathode (74).
An assembly according to anyone of claims 1 to 8 in which the main discharge is unballasted.
An assembly according to any one of the claims 6 to 8 further comprising means for varying the voltage across the pilot discharge and/or the main discharge to vary the intensity of the output of the lamp.