GIRAFFE: Calibration lamp stability
|
HC PLOTS |
Simultaneous ARC lamp: |
|
QC1 database (advanced users):
browse |
plot
|
Robotic flat field (FF) lamp: |
|
Nasmyth FF lamp: |
|
browse |
plot
|
ARC lamp: |
|
browse |
plot
|
The fluxes of all GIRAFFE calibration lamps are monitored here.
It is important to control the calibration lamp performance, both
over and under exposure should be avoided. Over-exposure is an
important issue for the flat-field lamps because the calibration files then
are useless (at least for removing the pixel-to-pixel gain variations).
Under-exposure for flat-field lamps results in compromised SNR,
and for arc lamps in poor emission line detection and low-quality dispersion
solution.
The calibration lamp fluxes are monitored through regular HC calibrations and deliver health check parameters.
They are monitored to discover aging effects,
filter degradation or problems with the gripper.
We monitor fluxes of two flat-field lamps and two arc lamps:
- the arclamp of the simultaneous calibration
unit:
used in operations only exceptionally (for high-precision wavelength
calibration), simultaneously with the science exposure;
used as daily health check to register
the spectral format;
- the flat field lamp on the fibre positioner
calibration unit ("robotic flat"):
used in operations as standard lamp
for flat-field exposures, taken as daytime calibration;
- the nasmyth lamp ("nasmyth flat"):
used for the IFU modes as night or twilight calibration since they provide a better illumination across the FOV.
- the arc lamp on the fibre positioner calibration
unit:
used in operations as standard wavelength calibration, taken as daytime
calibration;
Simultaneous arclamp
An exposure (about 120 sec integration time) with the arclamp
of the simultaneous calibration unit is daily taken. With the HR grating, they are taken for each of the five fibre systems.
For the LR grating, only Medusa1 is taken. See the caption for the exact settings used.
These data are used as check of the spectral format (motions of the 2 gratings), of the lamp stability and of the fibre systems health.
Find a description of the calibration files and of the grating stability here.
QC1_parameters
FITS key |
QC1 database: table, name |
definition |
class* |
HC_plot** |
more docu |
[calculated by the QC procedure] | giraffe_simlamp..flux | calculated flux of selected line, for lamp trending [ADU per sec] | HC | | |
*Class: KPI - instrument performance; HC - instrument health; CAL - calibration quality; ENG - engineering parameter
**There might be more than one, always check the link 'HC'. |
Trending
The trending
plot has two panels, one with the fluxes from the five HR settings, one with
the fluxes from the Medusa1 LR setting (all are taken as daily health checks).
|
Plots 1-5: Upper panel, fluxes of the simultaneous calibration arclamp in the five HR settings.
|
Panel 1 shows the lamp flux (in counts per sec) as measured in each
slit system with the HR setting (H525.8B).
It shows the lamp performance modulated by the fibre efficiencies. The flux difference
between the Medusa and the IFU/Argus systems is due to the different fibre diameter
(1.2 arcs versus 0.5 arcs).
|
Plot 6: Lower panel, the same for Medusa1 L543.1
|
The second panel shows the fluxes for the Medusa1 LR grating.
If a certain
trend effect is visible in *all* slits, it is likely to be caused by the
lamp. If it shows up in one fibre system only, it originates
in that particular fibre system.
Scoring&thresholds Simultaneous arclamp
Thresholds are set such that over-exposure is avoided. The risk of the lamp getting too bright is that
bright emission lines might affect the science signal of nearby fibres through cross-talk. This is
very much depending on setup (central wavelength), and the upper threshold is set conservatively and empirically. The
lower threshold is set such as to detect lamp failures safely. As the FULL plot shows, lamp fluxes have been
fairly fluctuating over the years, and also seem to have a seasonal component, all this arguing for a
loose and conservative thresholding.
History
The total lamp flux usually slowly increases with time. It is dimmed electronically
when it exceeds a certain threshold, to avoid saturation. This happened in 2004-02,
in 2007-03 and again in 2017-11. In the later case, PPRS-070892 collects the related activities. In 2005-07 ... 2005-09, the lamp faded and needed to be stabilized.
Algorithm Simultaneous arclamp
One emission line in one of the simultaneous calibration fibres is
selected. Its flux is integrated (in counts per sec). This
is done by a QC procedure since there is no pipeline recipe.
Flat field lamp (robotic)
The flux of the robotic flat-field lamp is measured in a pattern
involving the L543.1 setup in every slit system. A single flat frame is measured.
The lightpath between the lamp and the detector also includes interference
filters. Hence any effect observed in the efficiency trending could be due to either
the lamp or the filter.
The parameter giraffe_flat..flattype is either
ROB=robotic, or NAS=nasmyth. Here it is always ROB.
QC1_parameters
FITS key |
QC1 database: table, name |
definition |
class* |
HC_plot** |
more docu |
[calculated by QC procedure] | giraffe_flat..flux | mean signal = mean_raw/exptime [ADU/sec] | HC | | |
*Class: KPI - instrument performance; HC - instrument health; CAL - calibration quality; ENG - engineering parameter
**There might be more than one, always check the link 'HC'. |
Trending
Panels of the trending
plot:
|
Boxes 1-5: FF lamp efficiency. Data are from HC calibrations in the L543 setup, for all five fibre systems
|
The lower panel shows the same data, plotted together, to reveal any
slit-specific effects.
|
Box 6: Same data as above, all fibre
systems plotted together. In this plot, "coherent" changes in the signal can
be best seen, indicative of a lamp problem, as opposed to changes in a single
fibre system indicating a problem with that fibre system.
|
Plots of spectrally resolved relative fibre transmission can be found
here.
Scoring&thresholds Flat field lamp (robotic)
Upper thresholds both in the FFLAMP and the FFLEVEL reports are motivated by detecting saturation in the flat-fields.
The flux value for saturation (for an exposure time fixed in the templates) depends on the setup
(central wavelength) and can only loosely be defined in the HC settings. The lower threshold is set empirically
to detect lamp failures and compromised flat field quality resulting in insufficient SNR.
The QC1 values are averaged across the whole CCD and therefore include both illuminated regions (with the fibre signal) and
inter-fibre regions (without a signal).
History
Date |
event |
around 2007-10-01 | flux strongly decreased due to gripper problem; adjusted properly afterwards |
around 2008-01-14 | Low flux observed, due to bad positioning of the calibration mirror located inside of the OzPoz calibration unit |
Algorithm Flat field lamp (robotic)
The mean exposure level of the raw file across the whole file is measured (mean_raw) and then divided by the exposure time,
the result is the flux and comes in ADU/sec. For historical reasons the calculation is done by the QC procedure.
Nasmyth flat lamp
Argus nasmyth flats (flattype=NAS) are acquired everytime the standard
star efficiency monitoring is executed (every 1-2 months). Furthermore it may also
be triggered by an Argus SCIENCE OB and a standard star exposure. These
flats provide a better relative illumination than the robotic flats, so that the
relative fibre signal level is more evenly distributed.
Nasmyth flats also exist for IFU1 and IFU2 but are so sparsely distributed
over time and settings that a trending is impossible.
The parameter giraffe_flat..flattype is either
ROB=robotic, or NAS=nasmyth. Here it is always NAS.
QC1_parameters
FITS key |
QC1 database: table, name |
definition |
class* |
HC_plot** |
more docu |
[calculated by QC procedure] | giraffe_flat..flux | mean signal = mean_raw/exptime [ADU/sec] | HC | | |
*Class: KPI - instrument performance; HC - instrument health; CAL - calibration quality; ENG - engineering parameter
**There might be more than one, always check the link 'HC'. |
Trending
|
Box 1: Nasmyth lamp efficiency (Argus, L543.1)
|
Box 1 of the trending
plot displays the Argus L543.1 values. This setup is used by the Argus STD efficiency
monitoring, and also complies with the setup used for the robotic flat lamp monitoring.
|
Box 2: Nasmyth lamp efficiency (all settings). Blue marks
the L543.1 values, green the HR settings, red the LR settings.
|
Box 2 in the trending report collects all existing nasmyth lamp
data for Argus, to give an indication about the wavelength dependency of the lamp
efficiency.
History
Date |
event |
around 2007-02 | flux in the L543.1 setting increased, to provide better statistics |
Scoring&thresholds Nasmyth flat lamp
In principle the same arguments apply as for the robotic
flats. But the Nasmyth flats taken very rarely (with the
Nasmyth screen, costly in terms of nighttime or twilight time), as part of
Argus science OBs. Therefore
there are very few comparison points from the trending, so no scoring, and the thresholds are
very loosely defined.
Algorithm Nasmyth flat lamp
The average flux is obtained by dividing the average signal in the first
raw file by the EXPTIME.
ARC lamp
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.
QC1_parameters
FITS key |
QC1 database: table, name |
definition |
class* |
HC_plot** |
more docu |
lamp efficiency | giraffe_wave..lamp_effic | average signal of non-SIMCAL fibres (ADU/sec) | HC | | |
*Class: KPI - instrument performance; HC - instrument health; CAL - calibration quality; ENG - engineering parameter
**There might be more than one, always check the link 'HC'. |
Trending
Panels of the trending
plot:
|
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.
|
Boxes 6, 7: ARC-lamp efficiency (low-resolution setting L543.1).
|
Scoring&thresholds ARC lamp
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.
History
Date |
event |
2006-06-01 | ARC-LAMP flux re-adjusted to smaller values |
around 2007-10-01 | ARC-LAMP flux strongly decreased due to gripper problem; adjusted properly afterwards ; during that time the quality of the dispersion solution was reduced |
around 2008-01-14 | Low flux observed, due to bad positioning of the calibration mirror located inside of the OzPoz calibration unit |
Algorithm ARC lamp
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.
|