The performance of the robotic arc lamp is measured in a pattern involving two settings: the high-resolution H525.8B setup, and the low-resolution L543.1 setup. The former is measured in every slit system, the latter only in the two Medusa's. The pattern repeats every week. TITLE QC1 parameters TABLE_NAME qc1_parameter TABLE_TYPE QC1_TABLE TABLE_ROW1 lamp efficiency SEP giraffe_wave..lamp_effic TITLE Trending
Panels of the trending plot:
TABLE_NAME boxes15 TABLE_TYPE FIGURE_1B TABLE_IMG1 "../img/arclamp1.png" width="735" height="249" border="1" TABLE_CAP1 Boxes 1-5: ARC-lamp efficiency (high-resolution setting H525.8B).The lower panel shows the low-resolution Health Check setting, L543.1, for Medusa1 and Medusa2.
TABLE_NAME boxes67 TABLE_TYPE FIGURE_1 TABLE_WIDTH 80% TABLE_IMG1 "../img/arclamp2.png" width="317" height="202" border="1" TABLE_CAP1 Boxes 6, 7: ARC-lamp efficiency (low-resolution setting L543.1). TITLE Scoring&thresholds The upper thresholds are motivated by avoiding too high lamp fluxes. Although saturation is not as much an issue as for flat field lamps, the pipeline algorithm will reject saturated emission lines, and the number of accepted lines determines the wavelength calibration quality. With a lamp being too faint, more and more emission lines will be lost in the noise floor, again with the result of a compromised calibration quality. The ideal exposure level is close to the upper limit, but strongly depends on the setting (central wavelength).The thresholds of these HC settings are set to discover lamp flux issues in due time. TITLE History TABLE_NAME histo1 TABLE_TYPE HISTORY TABLE_ROW1 2006-06-01 SEP ARC-LAMP flux re-adjusted to smaller values TABLE_ROW2 around 2007-10-01 SEP ARC-LAMP flux strongly decreased due to gripper problem; adjusted properly afterwards ; during that time the quality of the dispersion solution was reduced TABLE_ROW3 around 2008-01-14 SEP Low flux observed, due to bad positioning of the calibration mirror located inside of the OzPoz calibration unit TABLE_END TITLE Algorithm The mean exposure level of all emission lines in the raw file is measured and normalized by the exposure time, the result comes in ADU/sec. Note that the signal from the SIMCAL fibres affects the average value as delivered by the pipeline. Therefore the final QC1 value is currently calculated using fibres #2...#10 only. This fibre range is free from SIMCAL signal.