Date:
Tuesday, January 26, 2016
Telescope:
CTIO Blanco 4-m
Submitter:
David James
Submitter Email:
Observer Support:
Manuel Hernandez
Night Assistant:
Claudio Aguilera
Also Present:
Katelyn Allers, Everett Schlawin, Nicole David
Problems:
Engineering for ARCoIRIS: Detailed comments below.
Comments:
Fourth and last night of our final Commissining/engineering run for the ARCoIRIS instrument.
This afternoon, Nicole David (assisted by Manuel Hernandez) conducted optical engineering tests in
the Cassegrain cage, testing different filters with the internal calibration unit. For four different
filters, we tested the colour balance of the Qz flat-fields as measured by ARCoIRIS. Results are as
follows [Spectral orders are 8-->3, left-to-right on the colour-balance plots].
Filter BG3: too little flux in spectral orders 4 & 5, although the bluest order (#8) now has enough
flux. This filter should not be used.
[see https://www.dropbox.com/s/geiuo2nvhfwv6mc/ColourBalance-Qz-BG3.jpg?dl=0]
Filter KC5: This filter seems to act as a neutral density filter in the sense that *very*, very little
flux was measured across the entire ARCoIRIS wavelength range. DO NOT USE.
[see https://www.dropbox.com/s/7cvff68604k1dk6/ColourBalance-Qz-KC5.jpg?dl=0]
Filter RG780: This filter did not seem to change the colour-balance of the Qz flat-field by very much
at all, and order #8 still contains very little flux.
[see https://www.dropbox.com/s/c5eltujeqolilli/ColourBalance-Qz-RG780.jpg?dl=0]
Filter KG1: The colour balance of this filter is very bad, with essentially no flux in spectral orders
5-8. These data are considerably worse than having no filter at all.
[see https://www.dropbox.com/s/m998nhmo55itir7/ColourBalance-Qz-KG1.jpg?dl=0]
In conclusion, none of these filters really had the desired effect of normalizing the colour balance
of the ARCoIRIS flat-fields with the Qz lamp. The domeflats that we are acquiring with the COSMOS
top-ring lamps are pretty good actually, where we can successfully trace the spectral orders and
flat-field the ARCoIRIS data. We would prefer a better system, but at the present time we can
successfully operate with the COSMOS lamp-system until an improved colour-balance solution for the
internal calibration unit can be found.
Omar E. and Rolo C. have today held discussions, distributed by email, regarding the Blanco TCS
definition of rotator angle, and its relationship with both the parallactic angle and the instrument
mechanical angle. They are implementing changes to the TCS to allow clearer definition of the
parallactic angle, allowing the telescope operators to directly enter parallactic angle as a rotator
angle driver.
Marco continues to work on his vacations, and today has greatly improved the ARCoIRIS control
environment, including security/safe mode features: His improvements, and their testing, include:
(a) "test" button on the Spectrograph Detector GUI, allowing the user to take a test spectrum, which
is displayed, and saved/over-written to /tmp/test.fits. We ran tests of this feature, and it works
well -- thanks MB!
(b) We tested the OBSTYPE=calibration function tonight, with some successes while some irregularities
remaining. For all comparison lamps, except for the Xenon lamp test, once the TOs power-on the lamps
through the TCS GUI, we find the following.
b-I): taking a single spectrum results in an arc lamp spectrum as expected, but the lamp stays on.
b-II): taking another single spectrum, the software turns the arc lamp off, and proceeds to take
an image (essentially a dark frame).
This is good -- however, we probably want to turn off the arc lamp immediately after the first
spectrum to allow enough time for the lamp to completely turn off and dim completely before taking
the lamps-off image.
We note that when we tested the Xenon lamp, the software changed behaviour. When we took a single
spectrum, by pressing "start" as normal, a first image was taken (which is a nice arc lamp spectrum),
followed by the lamp being turned off, and a second image being taken (a dark frame). So, a single
press of the "start" button resulted in two images ... an arc and a dark ... which is actually what
we want to have happen. **After** your vacation Marco, can you please investigate these different
behaviours ?
(c) The compass pointer of the SV Image Detector window **only** shows the correct orientation
(9-o'clock) for the 270-degree rotator angle. For rotator angles of 0-degrees, 90-degrees (or
very nearly, because the rotator cannot actually reach exactly 90-degrees) and 180-degrees, the
compass points to the wrong place. Marco, I'll make a schematic for you, showing how the compass
actually points as a function of rotation angle.
(d) Marco notes that he is quite-rightly deferring testing of various issues and requested control
environment upgrades detailed in last night's E-o-N report until the instrument is off-telescope
and back in the Coude room.
We note that the GS3 GMAP GUI on the pcguider machine has a couple of oddities.
(d-I) The UT time is 17-minutes behind that of the TCS reported UT time (which agrees with that
of the US Naval Observatory).
(d-II) The calculation of its Parallactic Angle is 10s of degrees different from the Parallactic
Angle that the ARCoIRIS TCS panel shows (which agrees with that of the JSkyCalc tool).
We note that one (only one) image from the previous night's data (25-Jan-2016), image SPEC_Eng0400,
contained null or blank TCS information across all FITS header fields. Did the TCS handshake fail,
or was this a different error internal to the TCS ?
We thoroughly tested the GUI_Tool_GRID feature "to to BB", which allows the user to slew from an
on-slit nod position to a Background Box, placed at some off-slit location on the SlitViewer
detector. This feature allows the user to acquire a sky-background spectrum for an extended source
for instance. Marco, this new feature works extremely well, with only one small error/failure.
For a simple one position nodding pattern, A --> BackBox, the observations worked well. The same
is true for two, three and four nodding positions; That is to say,
Nod A --> BackBox (return to A)
Nod A --> BackBox, Nod B --> BackBox (return to A)
Nod A --> BackBox, Nod B --> BackBox, Nod C --> BackBox (return to A)
Nod A --> BackBox, Nod B --> BackBox, Nod C --> BackBox, Nod D --> BackBox (return to A)
These tests were all successful.
However, when we used these four nod positions, and included a "repeats" entry of 3, the
observations all executed as expected, but on the very final step, the final BackBox exposure
was observed, but the star was *not* returned to position A. This was the only error/failure.
All observations were still executed, but the procedure did not quite finish.
Finally, using a rich star field (open cluster NGC 2527), we confirm the orientation of the SV
field as a function of rotator angle (see https://www.dropbox.com/s/kf85hkor8uv0c83/ARCoIRIS-onsky-OrientationJan2...).
All-in-all, a busy and thoroughly successful night !
saludos
DJJ, KNA, ES, ND
Program:
Target of Oportunity:
No
Time Observed:
0.0
Time Engineering:
9.0
Time lost technical:
0.0
Time lost weather:
0.0
Time lost other:
0.0
Description:
ARCoIRIS Engineering
Primary Investigator:
David James
Email:
Institution:
CTIO
Instrument:
CTIO 4m ArCoIRIS
Organization:
NOAO
Total Program Time:
9
| Total Times | Time Observed | Time Engineering | Time lost technical | Time lost weather | Time lost other | Total Program Time |
|---|---|---|---|---|---|---|
| 2016-01-26 | 0.0 | 9.0 | 0.0 | 0.0 | 0.0 | 9 |
Printer-friendly version