FR2871290A1 - Ultraviolet radiation emitting method for barrier excimer lamp, involves preparing bulb by configuring its wall to place working fluid in two paths that permit heating and cooling of fluid for leading to convection circulation of fluid - Google Patents

Abstract

The method involves preparing a bulb by configuring its wall for placing working fluid in two paths. Each path permits heating and cooling of the fluid for leading to a convection circulation of the fluid in the path when chosen emitting direction favors the path. The circulation assists cooling of the fluid. Two linear circuits are defined in the respective paths for the circulation, and extend in two crossed surfaces, respectively. An independent claim is also included for a barrier excimer lamp for implementing an ultraviolet radiation emitting method.

Description

An object of the present invention is a discharge barrier lamp.

  The principle of such a lamp is described in the "Discharge Handbook", Elektrogesellschaft, June 1989, 7th edition, page 263. Its radiation is generated by a dielectric working fluid subjected to electrical discharges. Typically this

  fluid is a gaseous medium at low pressure consisting of a rare gas and / or a halogen. Under the effect of a discharge, it forms excited species whose radiative electronic de-excitation transitions generate radiation to be emitted. These excited species are typically excimer or exciplex type molecules. The lamp then emits a particularly monochromatic ultraviolet radiation.

  The working fluid is confined in an ampoule whose walls are typically made of vitreous silica. These walls form two coaxial tubes constituting an inner tube and an outer tube, and this fluid is confined in the annular space between these two tubes.

  Electrical discharges are typically caused by steep high voltage pulses. Typically these pulses have a maximum voltage of several kilovolts and they last a few hundred nanoseconds and are repeated at a frequency of a few tens or hundreds of kilohertz. They are applied between on the one hand an internal electrode located in the inner tube of the bulb and connected and on the other hand an external electrode applied around the outer tube. The walls of these two tubes then constitute two dielectric discharge barriers. Only the internal electrode is brought to a high voltage.

  Such an ultraviolet lamp can be used for example in photochemistry or for industrial surface treatments, and also in medicine, especially in dermatological treatments such as psoriasis.

  Electrical discharges generated in the working fluid can excessively heat up this fluid. It is widely recognized that efficient cooling of this fluid is an essential condition of the longevity of the performance of such a lamp. This was confirmed by the work of the High Current Electronics Institute, a Siberian branch of the Russian Academy of Sciences, to which many of the present inventors belong. The difficulties in obtaining sufficient cooling increase with the power of the lamp and more particularly with the pfd of the radiative flux to be emitted.

  A first discharge barrier lamp is known from patent documents EP0517929 and CA2068574 (Von Arx). To ensure the necessary cooling these documents propose to immerse the bulb and the electrodes in a circulating cooling fluid preferably consisting of water. The choice of this fluid allows efficient cooling. But it requires to make troublesome and expensive arrangements. In particular, the water used must be kept at a high purity level to ensure proper electrical insulation. As for the means necessary to ensure the circulation of water and maintain the tightness of the circuits, they are heavy and bulky.

  A second discharge barrier lamp is known and is called lamp I by US6379024 (Kogure). The electrodes of this lamp have limited angular extents around the axis of the bulb so that only a portion of the working fluid is subjected to electric discharges. The necessary cooling is provided by water flowing in a conduit which extends into the inner tube of the bulb. Such a conduit may be used to electrically isolate the water from the inner electrode. But it then has the disadvantage of limiting the effectiveness of heat transfer to water. In addition, the means for ensuring the circulation of water and maintaining the tightness of the circuits in such a lamp are heavy and bulky. This may be why the document mentions the possibility of using air instead of water. But it is clear that the pfd of the radiative flux emitted by such a lamp should be strongly limited if this lamp was to be cooled in the air.

  A third discharge barrier lamp is known from US2004004422 (Falkenstein). A heat pipe extends between on the one hand a hot part engaged in the inner tube of the bulb of this lamp and on the other hand a cold part located and cooled outside this tube. This conduit is a sealed tube in which a liquid evaporates in the hot part, the vapor condenses in the cold part, and the condensed liquid returns to the hot, for example by capillarity, to evaporate again. It makes it possible to extract from the inner tube a particularly great thermal power. But it is effective only in a relatively narrow temperature range, and its effectiveness is limited by that of the transfer of heat from the working fluid to its hot part. In addition, a sufficiently powerful cooling system must be installed at its cold end and be compatible with correct electrical insulation.

  The present invention is intended in particular to make it possible: to increase the pfd of the radiative flux emitted by such a lamp, to limit the temperature of the working fluid so as to limit the decrease in lamp performance with service life, to limit the weight, the bulk and the cost of maintenance of the lamp, to make the lamp easily manipulated and, more particularly, portable, to facilitate a correct electrical insulation of an internal electrode subjected to high voltage pulses, and for this, to ensure a sufficiently powerful cooling of the working fluid by air.

  For these purposes it particularly relates to a radiation emission method in a controlled direction. This method includes the following steps: preparing a bulb having an at least partially dielectric wall and at least partially transparent for radiation, confining a working fluid in the bulb, installing this bulb and electrodes in a housing with arrangement of a window for an emission of this radiation from this bulb towards the outside of this housing, orientation of this housing to direct this window towards a chosen direction of emission, localized application of successive electrical discharges to the fluid working with these electrodes through dielectric regions of the wall of the bulb to cause the emission of this radiation by the fluid through this window, these discharges also causing a heating of the fluid, and cooling of this fluid through this wall during the application of electrical discharges, II is characterized by the fact that the preparation the bulb includes a configuration of its wall to develop for the working fluid two channels each adapted to allow the heating and cooling of the fluid to cause a convection flow of the latter in this way when the direction The selected transmission channel favors this route, this circulation being able to substantially assist this cooling, two mean linear circuits being respectively defined in these two channels for this circulation and respectively extending in two mutually crossed surfaces.

  With the aid of the attached schematic figures, examples will show how this invention can be implemented. When the same element or an element providing the same functions is represented in several of these figures, it is designated by the same letters and / or reference numerals.

  Figure 1 shows a view of a first lamp according to this invention in cross section relative to an axis of a bulb of this lamp, a spacer of this bulb is not shown.

  FIG. 2 represents a cross-sectional view on an enlarged scale of an inner tube of the ampoule of FIG. 1.

  Fig. 3 is a cross-sectional view of an inner tube of an ampoule of a second lamp according to this invention, with indication of the electrode positions of that lamp.

  FIG. 4 represents an axial sectional view of the bulb of the second lamp according to this invention, with indication of the electrode positions of this lamp. FIG. 5 represents a perspective view of a first convection circuit in a first position of the bulb of FIG. 3.

  FIG. 6 represents a perspective view of a second convection circuit in a second position of the bulb of FIG. 3.

  FIG. 7 represents a perspective view of the assembly of the two convection circuits of FIGS. 5 and 6.

  Fig. 8 is an enlarged cross-sectional view of an inner tube of the bulb of a third lamp according to this invention.

  Fig. 9 is an enlarged cross-sectional view of an inner tube of the bulb of a fourth lamp according to this invention.

  According to FIG. 1, a discharge barrier lamp emits ultraviolet radiation flux represented by two arrows such as arrow 1.

  Typically, it is formed in a housing 2 preferably made of metal or an internally metallised plastic. The light is emitted by an ampoule through a window F. This bulb is typically constituted by an inner tube T1 and a coaxial outer tube TO and consisting of a vitreous silica such as quartz sold under the reference GE214 or GE219 by the firm General Electric. Such a bulb is shown at TT in Figures 3 and 4. It confines a working fluid in the annular space between these two tubes. This fluid is typically a gas or gas mixture. The pressures of such gas mixtures are between 0.05 and 1 bar, preferably between 0.1 and 0.3 bar. For example, in a dermatological application, the mixture of gases will be composed of Xe and Cl2, in the volume proportion of 250/1, for a total pressure of 114 mmHg. The emitted radiation will then have a similar wavelength. 308 nm, and it will find application for the treatment of dermatoses such as psoriasis or vitiligo.

  An inner electrode El, also shown in Figure 2, and an outer electrode EO subject this working fluid to electrical discharges through the walls of these tubes. According to FIG. 1, the internal electrode El receives for this purpose high voltage pulses which are supplied by a generator 3, the external electrode EO being connected to the earth constituted by the housing 2. The voltage of these pulses is preferably between 1 and 15 kV, and more preferably between 7 and 11 kV and their repetition frequency is preferably between 30 and 150 kHz and more preferably between 70 and 110 kHz.

  The window F is constituted by a portion of the external electrode EO, only this part being transparent, or at least semi-transparent, for the emitted radiation.

  In the case of the lamp shown in Figure 1, this electrode surrounds the outer tube TO 360 degrees. In this case it preferably comprises two unrepresented metal layers. An inner layer is for example constituted by a sheet of aluminum or an alloy of Al and Mg 100 μm thick wound around the tube TO. This sheet has a width for example equal to the perimeter of the outer tube plus 1 to 5 mm to perform a slight recovery of the sheet. It has been cut beforehand to form the window F. A transparent layer is constituted by, for example, a wire wound helically with non-contiguous turns around the inner layer, and clamped on this inner layer. This wire is for example a nichrome wire 0.1 mm in diameter and is wound with a pitch of about 0.7 to 1 mm between each turn. The inner layer and this wire are held in contact with the tube TO by two circular flanges. This inner layer and these flanges are not shown and the portion of this wire which constitutes the window F is symbolically represented in FIG. 1 by a dashed line.

  The external electrode may further comprise a plate EO which is shown in Figure let which maintains the bulb in the housing 2 through two extensions such as 6 which join the wall of this housing. The internal surfaces of the two electrodes and these extensions are processed to reflect the radiation emitted by the bulb so as to reinforce and uniform the flux emitted by the lamp.

  The heat from the electric discharges is evacuated by the air flowing in the tube T1 and around the external electrode E0. This air is driven by one or preferably two fans such as 4 arranged at both ends of the TT bulb. It enters the housing 2 and exits through openings such as formed in the walls of this housing.

  Figures 3 and 4 show the axis LA of the TT bulb and the angular extents Al, AO, and AF and longitudinal LI, LO and LF electrodes El and EO and the window F, respectively. The dimensions of the bulb and these areas are chosen according to the use envisaged for the lamp. The same is true of the composition and the pressure of the working fluid and the characteristics of the pulses supplied by the generator 3.

  The length LC of the internal VC space at the bulb TT is preferably between 10 and 2000 mm, and more preferably between 100 and 200 mm. The diameter of the inner tube T1 is preferably between 10 and 50 mm, and preferably still close to 20 mm. The diameter of the outer tube TO is preferably between 20 and 100 mm, and preferably still close to 43 mm. The thickness of these tubes is preferably between 1 and 3 mm, and more preferably close to 1.5 mm. And the distance between the outer surface of the inner tube and the inner surface of the outer tube is preferably between 5 and 25 mm, and more preferably close to 10 mm.

  A lamp according to this invention may have various shapes and arrangements. More generally than above, and in the context of a first group of advantageous arrangements, it comprises essential elements which are shown in the figures by way of example and which are as follows: a wall TI, TO confining the working fluid in a confinement space VC. This wall is at least partially dielectric and at least partially transparent for the radiation.

  Two electrodes El and EO arranged outside the bulb and extending mutually opposite on either side of a fraction of the confinement space. This fraction constitutes a discharge space VD. In the case where only a fraction of the extent of one of these two electrodes is opposite the other electrode, the discharge space is limited to this fraction of extent. That is, the extent of the discharge space is the extent common to both electrodes. Those areas of the wall that are located between these two electrodes on either side of this discharge space are dielectric. They constitute respectively two barriers of discharge.

  Means 3 for applying between the two electrodes E1 and E0 an alternating voltage electrical adapted to induce the necessary electric discharges in the fraction of the working fluid that is present in the discharge space.

  Finally means 4 for cooling the working fluid through the wall of the bulb.

  According to FIGS. 5 to 7 and in the context of the first group of advantageous arrangements, the wall of the bulb forms for the working fluid at least a first and a second circulation lanes W1 and W2 having a common part constituted by the discharge space. Each of these channels is capable of channeling a circulation of this fluid. It passes for this by a space looping this way and constituting respectively a first BI and a second B2 looping spaces. It offers the molecules of the working fluid a multitude of possible paths. In this multitude, an average path defines for the circulation of this fluid in this way a closed average linear circuit. The two such circuits respectively constitute a first and a second convection circuit. These two channels can not be accurately represented, they are represented in the form of these circuits W1 and W2. The first and second convection circuits extend respectively into first and second mutually crossed loopback surfaces P1 and P2. These surfaces are typically substantially planar and mutually perpendicular and the two loopback spaces typically have a second common portion PC away from the discharge gap VD.

  Each of the two circulation lanes W1 and W2 has a passage section for the fluid at each point of the convection circuit associated with this lane. This section has an area and all areas of the passage sections of that route include a minimum area and an average area. This minimum area is preferably greater than 30% of this average area.

  The lamp is preferably steerable in several directions so that the circulation of the working fluid is established preferably in one or the other of the two traffic lanes W1 and W2 in the direction of orientation of this lamp. This circulation is a convection circulation. It is caused by the heating of this fluid in the discharge space VD and by its cooling in the looping space B1 or B2.

  Some of the areas of the bulb wall preferably surround the VC containment space and form TO peripheral wall areas. Some other of these zones form an internal channel C1 surrounded by this confinement space, these other zones constituting inner wall zones TI. This channel has two ends C1 and C2 and an axial line LA extending between these two ends. It has a cross section at each point of this length and this cross section has an area and a perimeter. Still others of these wall areas connect these peripheral wall areas to these inner wall areas and form connecting wall areas. Lengths LI, LO and LF and two mutually opposite longitudinal directions are defined along this axial line. Two mutually opposite transverse directions VD PC and PC VD are defined with respect to this line. Al, AO and AF perimeter ranges are defined around this line.

  An electrode is disposed in this channel in contact with at least one inner wall zone and constitutes an internal electrode El. The other electrode extends in contact with at least one peripheral wall zone and constitutes an external electrode E0.

  According to FIG. 3, the discharge space VD has a perimetric extent AI substantially less than one complete revolution, so that a remaining part of this turn constitutes the first looping space BI and that the first convection circuit W1 is extends over the whole of this round around the internal channel.

  According to FIG. 5, this circuit is active, that is to say that the first channel W1 is the seat of a convective circulation, when the axis of the bulb is approximately horizontal, provided of course that the window F, under which occur the discharges and therefore the heating, is not directed upwards. This circuit extends in a vertical plane perpendicular to the axis of the bulb. The circulation extends with a single direction of rotation but with a gradually decreasing speed, until the vicinity of the ends of the tubes T1 and TO.

  According to FIG. 4, the discharge space has a length LI extending between two longitudinal ends D1 and D2 of this space and the confinement space has a length LC extending between two longitudinal ends C1 and C2 of this space. space.

  In the context of the first group of advantageous arrangements, a gap extends between each of the two longitudinal ends of the discharge space and the closest of the two longitudinal ends of the confinement space. The two such intervals constitute a loopback interval Cl D1 and a loopback interval D2 C2. The second convection circuit W2 then includes in succession from this discharge space: a first segment S1 extending in a first longitudinal direction C1 C2 and constituted by a first looping interval, a second segment S2 consisting of two branches 2R , 2L extending in parallel on either side of the internal channel according to an average of a first transverse direction VD, PC in the length of the first looping interval, a third segment S3 extending along the second longitudinal direction C2 Cl in a fraction of the confinement space transversely opposed to the discharge space, a longitudinally median portion of this transversely opposite fraction constituting a common portion PC at the first and second looping spaces, a fourth segment S4 consisting of two branches 4R, 4L extending in parallel on either side of the internal channel in an average of the second transverse direction PC, VD in the length of a second loopback interval, a fifth segment S5 extending in the first longitudinal direction and constituted by the second loopback interval, and a sixth segment S6 extending in the first longitudinal direction in space discharge.

  According to FIG. 6, this circuit is active, that is to say that the second path W2 is the seat of a convective circulation, when the axis of the bulb is approximately vertical, and that whatever then is the position of the window F. FIG. 7 schematizes the relative positions of the circuits W1 and W2 with respect to the bulb. The axis of this one is supposed to be vertical like the segment S3. That is, compared to the position of Figure 5, the bulb is assumed to have tilted 90 degrees. In the position of Figure 7 only the circuit W2 is active. It extends in an axial plane, thus vertical P2. The circuit W1 is only virtual. Its plane was vertical in Figure 5 but the tilting of the bulb makes it appear in Figure 7 as extending in a horizontal plane P1. These two planes intersect in a virtual straight line passing through a central point of the discharge space VD and a central point of the common part PC. They are mutually perpendicular.

  Each loopback interval preferably has a length LB greater than 15% and more preferably 20% of the length LC of the confinement space VC.

  The axial line LA is preferably rectilinear. It then constitutes an axis of the TT bulb. The peripheral wall zones and internal wall zones constitute respectively an outer tube TO and an inner tube T1. These two tubes are typically transparent and dielectric over their entire surface. They are for example cylindrical and coaxial, the previously mentioned perimeter tracts then being angular extents. They have common longitudinal ends C2 and Cl, it being understood that one and / or the other of these tubes may have one or more extensions which have been for example useful for the realization of the enclosure, but which do not participate confining the working fluid. Such extensions of the tube T1 appear in FIG. 4. At least one of the electrodes E0 and E1 ends longitudinally at distances from these ends to constitute the looping intervals C1D1 and C2D2.

  The window F is defined by a diaphragm. It is constituted by the opening of this diaphragm. Its longitudinal and angular dimensions are advantageously less than those of the discharge space VD so that only a central fraction, and preferably a majority, of the flux emitted by the working fluid is transmitted to an external target through this window. The stream received by this target can then be homogeneous, which is useful in many applications, whereas it would not be if it consisted of the entire flow emitted by the working fluid.

  The external electrode EO is advantageously present in the emission window F. It is then transparent, at least partially, for the radiation of the lamp. It is opaque around this window to constitute the diaphragm that defines this window. Its longitudinal and angular ranges are then greater than those of the internal electrode El so that it is the latter which defines the extent of the discharge space VD.

  The angular extent AI of the discharge space VD is preferably between 5 and 180 degrees, and more preferably between 90 and 180 degrees. Its longitudinal extent LI is preferably between 60% and 70% of the length LC of the confinement space.

  The angular AF and longitudinal LF extents of the window F are preferably between 70% and 90%, and more preferably between 80% and 90% of those AD and LD of the VD discharge space.

  The angular AO and longitudinal LO extents of the outer electrode E0 are preferably between 110% and 130%, and more preferably between 110% and 120% of those AD and LD of the VD discharge space.

  FIGS. 2, 8 and 9 show a second group of advantageous arrangements which find application in the typical case where the wall TO, TI of the bulb TT forms the internal channel CI and its axial line LA, where one of the two electrodes extend in this channel in contact with at least one inner wall zone and constitute an internal electrode El, and where this electrode is made of a thermally conductive metal, this wall area being dielectric. In the context of this second group, the lamp further comprises a heat evacuator EV extending transversely in the internal channel CI while remaining at a distance from the inner wall areas TI. This evacuator is also made of a thermally conductive metal and it is in continuity at least thermal with the internal electrode so as to transmit heat transversely of this electrode to the cooling fluid. It extends longitudinally over at least a majority fraction, and preferably over at least the whole of a longitudinal extent common to both electrodes. As shown in FIGS. 8 and 9, it has a thermal contact area with the cooling fluid of at least 200% of the contact area of the internal electrode with the inner wall areas T1.

  In the case where, in addition, the inner electrode El has in each of the cross sections of the internal channel C1 a perimetric area substantially smaller than the perimeter of this section, the lamp preferably also comprises at least one dielectric spacer ET based on internal wall areas remote from the inner electrode to maintain this electrode and / or the heat evacuator EV.This spacer is for example made of mica and it was glued to the inner tube after the introduction of the internal electrode and the heat evacuator.

  According to Figures 2 and 8, the inner electrode El and the evacuator EV are formed by the same piece of metal. According to Figure 2, this part is a longitudinally extending tube. The section of this tube comprises on the one hand an arc constituting the electrode El and on the other hand a rectilinear, convex or concave segment constituting the evacuator EV. According to Figure 8, the metal piece El, EV is a folded sheet with longitudinal fold lines. Folding is carried out so as to form contacts between the plies and the electrode and possibly unrepresented contacts between the successive folds.

  According to Figure 9, the heat evacuator EV is formed by a plurality of tubes such as EV1 and EV2 which extend longitudinally in mutual transverse contact. Tubes such as EV1 have a larger diameter than tubes such as EV2. The diameters and the number of these tubes are chosen to form a large number of contacts between the tubes and between the tubes and the electrode El, and for the spacer ET to ensure a permanent support force in most of the zones. contact.

  The advantageous arrangements indicated above make it possible to obtain efficient air cooling and thus to avoid the heaviness and bulk of water cooling. They make it possible to realize a portable and orientable lamp emitting a pfd remaining greater than 60 mWlcm2 during more than 2000 hours.

Claims (10)

  1 - Method of radiation emission in a controlled direction (1), this method including the following steps: preparation of a bulb (TT) having a wall (TI, TO) at least partially dielectric and at least partially transparent for a radiation, this wall having zones, confinement of a working fluid in said bulb, installation of said bulb and electrodes (El, EO) in a housing (2) with arrangement of a window (F) for an emission said radiation from said bulb towards the outside of said housing, orientation of said housing for orienting said window towards a chosen transmission direction, localized application of successive electrical discharges to said working fluid using said electrodes through dielectric zones of said wall to cause an emission of said radiation by said fluid through said window, these discharges also causing a heating of this fluid, and re cooling of said working fluid through said wall during said application of electric discharges, characterized in that said preparation of a bulb includes a configuration of said wall for arranging for said two-way working fluid (W1, W2 ) each adapted to allow said heating and cooling to cause a convective circulation of this fluid in this way when said chosen transmission direction favors this way, this circulation being able to substantially help this cooling, two mean linear circuits being defined respectively in these two paths for this circulation and respectively extending in two mutually crossed surfaces (P1, P2).   2. The method of claim 1, wherein said cooling of the working fluid is provided by a flow of air.   3 - Lamp discharge barriers, this lamp comprising, a working fluid adapted to receive a succession of electric discharges and respond to each of these discharges by emitting a useful radiation while undergoing a parasitic heating, a bulb (TT) having a wall (TI, TO) confining said working fluid in a confinement space (VC), this wall (TI, TO) being at least partially dielectric and at least partially transparent for said radiation, this wall having zones, two electrodes (El, EO) extending out of said bulb and being mutually opposite each other on a fraction of said confinement space, this fraction constituting a discharge space (VD), those of said zones of the wall which are located between these two electrodes on either side of this discharge space being dielectric and respectively constituting two discharge barriers, means (3) for applying between the two said electrodes an alternating voltage electrical voltage adapted to induce said electric discharges in the fraction of said working fluid present in said discharge space, and means (4) for cooling said working fluid through said wall, characterized by the fact that said wall of the bulb forms for said working fluid at least a first (W1) and a second (W2) circulation lanes having a common portion constituted by said discharge space and each able to channel a circulation of this fluid passing through a space looping this path and respectively constituting a first (B1) and a second (B2) looping spaces, each of these channels defining for this circulation a closed average linear circuit associated with this path and respectively constituting a first (W1) and a second (W2) convection circuit, these first and second convection circuits respectively extending into a first (P1) and a second (P2) mutually crossed looping surfaces.   4 - lamp according to claim 3, wherein said means for cooling the working fluid are means (4) for circulating air in thermal contact with said wall (TI, TO).   - Lamp according to claim 3, wherein said first (P1) and second (P2) loopback surfaces are substantially planar and mutually perpendicular.   6 - Lamp according to claim 3, wherein said first (B1) and second (B2) looping spaces have a common portion (PC) away from said discharge space (VD).   7 - lamp according to claim 3, wherein each of said first (W1) and second (W2) traffic lanes has a passage section for said fluid at each point of said convection circuit associated with this path, this section having a area, all the sections of passage sections of this route including a minimum area and an average area, this minimum area being greater than 30% of this average area.   8 - Lamp according to claim 7, this lamp being steerable in several directions so that said circulation of the working fluid is established preferably in one or other of the two said traffic lanes (W1, W2) according to the direction of orientation of this lamp, this circulation being a convection circulation caused by a heating and cooling of the fluid respectively in said discharge space (VD) and in said looping space (B1, B2).   9 - Lamp according to any one of claims 3 to 8, wherein some of said zones of the wall of the bulb surround said confinement space (VO) and constitute peripheral wall zones (TO), some other of these zones forming an internal channel (C1) surrounded by this confinement space, these other zones constituting internal wall zones (TI), this channel having two ends (C1, C2) and having an axial line (LA) extending between these two ends, lengths (LI, LO, LF) and two mutually opposite longitudinal directions (C1 C2, C2 C1) being defined along this axial line, two mutually opposite transverse directions (VD PC, PC VD) being defined with respect to this axial line, and perimeter ranges (Al, AO, AF) being defined around this axial line, a said electrode in this channel in contact with at least one said internal wall zone and constituting an internal electrode (El), the other said electrode extending in contact with at least one said peripheral wall zone and constituting an external electrode (EO), said discharge space (VD) having a perimetric extent (Al) substantially less than one complete revolution, so that a remaining part of this turn constitutes said first looping space (B1) and that said first convection circuit (W1) extends over the whole of this turn around said internal channel, this discharge space having a length (LI) extending between two longitudinal ends (D1, D2) of this space, this confinement space having a length (LC) extending between two longitudinal ends (C1, C2) of this space, this lamp being characterized by the fact that an interval extends between each of said two longitudinal ends of the discharge space and the closest of the two said longitudinal ends of the confinement space, this gap constituting a looping interval (Cl D1, D2 C2) such that said second convection circuit (W2) includes in succession from this discharge space: a first segment (Si) extending along a first said longitudinal direction (C1 C2) and constituted by a first said loopback interval, a second segment (S2) constituted by two branches (2R, 2L) extending in parallel on either side of said internal channel according to an average first said transverse direction (VD, PC) in the length said first looping interval, a third segment (S3) extending along the second said longitudinal direction (C2 Cl) in a fraction of said confinement space transversely opposite said discharge space, (a longitudinally median portion of this fraction transversely opposed is a common part (PC) to the first and second looping spaces) a fourth segment (S4) consisting of two branches (4R, 4L) extending in parallel on either side of said internal anal on average said second said transverse direction (PC, VD) in the length of a second said looping interval, a fifth segment (S5) extending along said first said longitudinal direction and constituted by said second looping interval and a sixth segment (S6) extending in said first longitudinal direction in said discharge space.   - Lamp according to claim 9, wherein each said loopback interval has a said length (LB) greater than 15% and preferably 20% of said length (LC) of the confinement space (VC).   11 - Lamp according to any one of claims 9 and 10, wherein said axial line is rectilinear and constitutes an axis (LA) of said bulb (TT), said peripheral wall areas and inner wall areas respectively constituting a tube external (TO) and an inner tube (TI), these two tubes being transparent, dielectric, cylindrical and coaxial and having common longitudinal ends (C1, C2), said external electrode (EO) being transparent at least in said window emission (F), at least one (El) of said electrodes (EO, El) terminating longitudinally at distances from these ends to form said loopback intervals (Cl D1, D2 C2).   12 - lamp according to any one of claims 3 to 11, wherein a rare gas is mainly a said working fluid wherein said electric discharges can create excimer or exciplex emitting ultraviolet radiation.