VIR Calibration Lamps All of the material presented here has been extracted from the following reference: Melchiorri, R., G. Piccioni, A. Mazzoni, VIRTIS-M flight lamps, Rev. of Sci. Instr., 74(8), 3796-3801, doi:10.1063/1.1593784, 2003." [MELCHIORRIETAL2003] provides a comprehensive description of the calibration lamp design, testing, and spectral characteristics that will not be addressed here. The purpose of this document is to provide access to a small part of the material contained in the paper as documentation for the Dawn VIR instrument to those that do not have easy access to the journal article. Since the precise knowledge of the source is fundamental to the quality of the calibration, internal sources have the advantage that they may be characterized in a well controlled laboratory environment. For the VIRTIS-M spectrometer two lamps have been designed as internal sources, "visible" (VIS) and "infrared" (IR). Instead of using a commercial model, the lamps were custom built in order to control their characteristics. Each lamp has a tungsten filament. Tungsten was selected becuase it is easy to work with, has a stable optical emissivity, and a long lifetime. Each filament has been made by cutting a longer one at the nominal length (in order to have the designed electrical resistance). Typical filament dimensions are: 62.3 +/- 0.1 mm length, and 30 microns in diameter for a total number of eight spires per lamp [MELCHIORRIETAL2003]. The spectrum of the radiation emitted by the tungsten filaments is featureless. In order to allow a precise frequency calibration, spectral features have been added by inserting a holmium filter for the VIS model and a polystyrene filter for the IR model, whose features are detectable by VIRTIS-M (see Fig. 3, [MELCHIORRIETAL2003]). Figure 3: Holmium and polystyrene transmittance in the VIS-NIR region (modified after [MELCHIORRIETAL2003]) - provided as MELCHIORRI-FIG3.JPG Xenon gas is used in the lamps becuase this gas was found to be more stable than Argon during testing [MELCHIORRIETAL2003]. The pressure of this gas is around 110 hPa at 80 deg C. The use of Xenon may introduce problems at temperatures lower than the xenon boiling temperature (about -108 deg C). During the cold tests the lamps were cooled down the lamps to an ambient temperature of about -140 deg C where gas liquefaction may occur; however, this situation does not introduce substantial changes in the response of the lamp [MELCHIORRIETAL2003]. Presumably, the liquid xenon almost instantly evaporates as the filament is turned on. The minimum time needed for the filament to reach both thermal equilibrium and spectral stability is the 1st parameter to be determined [MELCHIORRIETAL2003]. During testing, a spectrometer provided a full spectrum of the lamp every 5 min, shown in Fig. 9 in [MELCHIORRIETAL2003]. It was assumed that 5 min is the minimum time for the filament to reach thermal equilibrium. Fig. 9 shows that the spectrum of the Xenon lamp was stable while the Argon lamp was not. This justifies the selection of xenon lamps, even if argon would have lower condensation temperature [MELCHIORRIETAL2003]. As shown in Fig. 9 (the right side), argon requires much longer times to acquire stability. FIG. 9. Each 5 min a spectrum is acquired. No difference is detectable for the xenon lamp (left), but argon presents spectral features changing in time (right, modified after [MELCHIORRIETAL2003]). see MELCHIORRI-FIG9.JPG The final lamp spectrum is a combination of the blackbody-like emission by the filament and the absorption features due to the transmission of the filters (see Fig. 17, [MELCHIORRIETAL2003]). Once the filter features have been subtracted, it is possible to fit the lamp spectrum with a Planckian curve, in order to determine the equivalent filament temperature (Fig. 10, [MELCHIORRIETAL2003]). This test gave the result T = 2600+/-100K for VIS and T=2400+/-100K for IR. FIG. 10. VIS lamp fit with an equivalent BB curve. Temperature estimated of ~2610 K (modified after [MELCHIORRIETAL2003]). see MELCHIORRI-FIG10.JPG The minimum time for thermal stability was found to be of about 5 min, which means that during the lifetime (and in the working conditions) the lamps will never be in equilibrium [MELCHIORRIETAL2003]. The voltage, current, flux, and temperature are measured. As previously, a mass spectrometer is used to control possible leaks. At the end of the each sequence, the monochromator acquires the lamp spectrum. The behavior of the filament resistance with the temperature has been monitored continuously. Any abrupt change would indicate some spurious contact among the spires. A fit to the curve (Fig. 15, [MELCHIORRIETAL2003]) provides the parameters in Table III. It is clear that for temperatures higher than -50 deg C the curve tends to a straight line. FIG. 15. Dependence of the VIS and IR filament resistance on the temperature; a linear fit for the VIS type gives b = 0.0151(Ohm/degC) and a =3.94 Ohm and for the IR type: b = 0.0151(Ohm/degC) and a =4.06 Ohm (modified after [MELCHIORRIETAL2003]). see MELCHIORRI-FIG15.JPG Table III Parameters of the 3rd order fit =============================================================== term VIS IR unit --------------------------------------------------------------- a 3.9692952 4.0696665 Ohm b 0.014799736 0.014292713 Ohm/degC c 1.9703929e-005 7.7951323e-006 Ohm/(degC)^2 d -7.7914420e-009 -2.199964e-009 Ohm/(degC)^3 =============================================================== FIG. 17. Spectral variations for IR and VIS lamps. On the left side it is possible to see the holmium features. On the right side pyroxene features are not present in this band region. The maximum of the Planckian curve does not change form cycle to cycle (modified after [MELCHIORRIETAL2003]). see MELCHIORRI-FIG17.JPG Resistivity and temperature are related to each other, in first approximation, by a linear equation. Assuming that geometrical factors do not change with temperature (in a first approximation), it is possible to evaluate the filament temperature by the knowledge of the resistance at ambient temperature [Eq. (1)]: R(T) 5.5 ------ + 5.0748 R(20K) Eq(1) T(K) = ----------------------- 0.0351286 In flight, there will be no possibility to analyze the lamp spectrum. Equation (1) will allow us to retrieve the spectral calibration. Table IV shows a comparison between the equivalent temperature measured by the electrical parameters and by the blackbody (BB) curve fit. TABLE IV. Equivalent filament temperature measured by electrical parameters and by BB curve fit (ep stands for electric parameters, and e stands for emissivity). Table IV ======================================================== Lamp model ep (K) e (K) -------------------------------------------------------- IR 2393 2436 VIS 2590 2576 ========================================================