Investigation of the gas-phase emitter effect of dysprosium in ceramic metal halide lamps

Abstract

The behaviour of electrodes operated with ac-currents in ceramic metal halide lamps containing Hg + NaTlDy iodide has been investigated experimentally. Using transparent YAG lamp tubes with the so-called Bochum model lamp as an outer bulb phase resolved measurements were performed of the electrode temperature and Dy density in dependence on the cold-spot temperature of the salt filling. The electrode tip temperature and electrode power loss are deduced from the temperature profile measured along the electrode axis. The Dy density in front of the electrode is determined by spatially resolved spectroscopic measurements of absolute line intensities. It is found that doping of a mercury lamp only with Dy iodide generates at low operation frequencies a pronounced emitter effect at the cathode but it declines with increasing frequency. In a lamp doped with NaTlDy iodide the formation and movement of Dy ions are hampered by Na ions accumulated in front of the cathode due to cataphoresis. As a consequence the lowering of the power loss by Dy is in part diminished. It is shown that a gas-phase emitter effect of Dy is effective for standard operation conditions of lamps in spite of the counteracting effects of a long time constant of the emitter effect and Na accumulation.

Investigating the gas phase emitter effect of caesium and cerium in ceramic metal halide lamps in dependence on the operating frequency

“…A plasma temperature of T pl ≤ 7000 K with an error margin of about 4 % was found within YAG lamps with similar parameters using the two Hg-lines [24,26].…”

“…For this purpose the image of the upper lamp electrode is adjusted perpendicular to the slit by turning the Dove prism through 45 • around the optical axis. A more detailed description of the measuring system for emission spectroscopy can be found in [14,24,26,33].…”

“…By an inverse Abel transformation of I L (y) radial emission coefficients L (r) are obtained if the discharge is rotational symmetric. Table 4: Spectroscopic data for the calculation of the plasma temperature T pl from Hg-line intensities [38] At first the plasma temperature T pl inside a YAG lamp is determined using two overlapping mercury lines at λ 1 = 576.96 nm and λ 2 = 579.07 nm [15,24,37]. The corresponding spectroscopic data is given in table 4 with the transition probabilities A 1 and A 2 , the statistical weights g 1 and g 2 and the excitation energies of the upper line levels E 1 and E 2 .…”

“…This situation seems to be present in a YAG lamp being seeded with DyI 3 and different other additives as NaI and T lI [22]. If HID lamps, which are seeded with T hI 4 [23], DyI 3 [24,25] or other metal iodides are operated with an ac or switched-dc current, it could not be decided definitely, whether the emitter effect is limited to the cathodic phase or also being effective within the anodic phase. To clarify this point, the gas phase emitter effect of Dy is investigated in a preceding paper [26] by spectroscopic measurements in front of the electrode and pyrometric measurements along the electrode rod in dependence on the operating frequency.…”

Why does the lumen maintenance of sodium–scandium metal halide lamps improve by VHF operation?

“…The supply of emitter to the electrode surface is via the gas phase (the gas-phase emitter mechanism [5,6]) and by diffusion of thorium from the inside of the electrode to the surface. Part of the emitter flux from the gas phase to the cathode surface is due to cataphoresis of the emitter ions and part is normal adsorption (see figure 9).…”