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.

To cite this version:J Reinelt, M Westermeier, C Ruhrmann, A Bergner, G M J F Luijks, et al.. Investigating the influence of the operating frequency on the gas phase emitter effect of dysprosium in ceramic metal halide lamps. Journal of Physics D: Applied Physics, IOP Publishing, 2011, 44 (22) Abstract. The dependence of the gas phase emitter effect of Dy on a variation of the operating frequency between some Hz and 2 kHz is investigated in a HID lamp. The buffer gas of the lamp consisting of Ar, Kr and predominantly Hg is seeded with DyI 3 , its burner vessel is formed from transparent YAG material. Phase and spatial resolved emission spectroscopy in front of the lamp electrode and pyrometric temperature measurements along the tungsten electrode are performed with a spectroscopic set up. Dy atom and ion densities in front of the electrode are deduced from absolute intensities of optically thin Dy lines and a plasma temperature, derived from the absolute intensity of mercury lines. Phase resolved values of the electrode tip temperature T tip and input power P in are obtained from temperature distributions along the electrode. Distinctly higher Dy ion and atom densities are measured in front of the electrode within the cathodic phase. With increasing operating frequency a reduction of both, atoms and ions, is observed in front of the cathode. In contrast, an increase of the ion density in front of the anode is seen. Moreover, the Dy ion density is drastically reduced by an additional seeding of the lamp with T lI. It is found that an up rating of the Dy ion density is correlated with a decline of T tip and P in . At higher frequencies this effect takes place not only within the cathodic phase but also within the anodic phase. The reduction of the average electrode tip temperature on the order of several hundred Kelvin compared to a YAG lamp with a pure mercury filling is explained by a Dy monolayer on the electrode surface which is sustained by a Dy ion current.Submitted to: J. Phys. D: Appl. Phys.Investigating the influence of the operating frequency on the gas phase emitter effect of Dy2

Abstract. Phase resolved temperature distributions are determined along a rod shaped tungsten electrode, by which an ac arc is operated within a model lamp filled with argon. Switched dc and sinusoidal currents are applied with amplitudes of several amps and operating frequencies being varied between 10 Hz and 10 kHz. The temperature is deduced from the grey body radiation of the electrode being recorded with a spectroscopic measuring system. Phase resolved values of the electrode tip temperature T tip and of the power input P in are determined comparing the measured temperature distributions with the integral of the one dimensional heat balance with these parameters as integration constants. They are supplemented by phase resolved measurements of the sum of cathode and anode fall called electrode sheath voltage. If a switched dc current is applied it is found that both quantities are within the cathodic phase only marginally higher than for a cathode being operated with a dc current. T tip and P in start to decrease for low currents and to increase for high currents at the beginning of the anodic phase. But with increasing operating frequency the deviations from the cathodic phase are reduced until they cannot be resolved for frequencies of several kHz. A more pronounced modulation but the same tendencies are observed with a sinusoidal current waveform. For 10 kHz a diffuse arc attachment with an almost phase independent electrode tip temperature, which deviates only marginally from that of a dc cathode, and an electrode sheath voltage proportional to the arc current is established with both current waveforms.

At cathodes of high intensity discharges two different modes of arc attachment are observed: a diffuse mode characterized by a low current density and a uniformly distributed high global temperature, and a spot mode with a high current density and a high local but low global surface temperature. For dc and low frequency operation an investigation of these different modes has already been presented. Different modes of arc attachment are shown in the case of RF operation (up to 1 MHz operation frequency) of the Bochum model lamp. Images of the arc attachment demonstrate that with increasing frequencies the occurrence of the spot mode is reduced. But it cannot be excluded that it arises accidentally at high frequencies. It is found that with increasing frequency the differences between the electrode tip temperature in the cathodic and the anodic half period are levelled and that the course of the electrode sheath voltage (ESV(t)) adjusts to the sinusoidal current waveform resulting in an increase in the average power input into the electrodes with proceeding adjustment. Moreover, it indicates a reshaping of the electrode boundary layers.

No abstract