MODS1 Dark Commissioning and Spring SV Run: UTC 2011 May 31
Observer(s): R. Pogge (OSU), O. Kuhn (LBTO)
Telescope Operator: G. Bechetti (LBTO)
Telescope Support: M. Midkiff (LBTO)
Instrument Support: D. Officer (LBTO)
Observed all night. Seeing was 0.5-1.0 arcsec, deteriorating towards the end. Started out hazy but mostly clear, and ended up with
high clouds coming in over use the last third of the night.
The instrument did well, a few glitches mostly pilot error. There were a couple of issues with the GCS and active optics, as noted
below, that slowed us down a great deal at the start of the night until we got things sorted out.
A good start for science verification observing, it will give people a taste of what we can do. Too bad the telescope and conditions
didn't always cooperate.
1345 MST: winds calm (under 10m/s), hazy and some clouds off way in the south. First bit of promising weather we've had in a while.
Setup a program based on the highest priority SV targets from the partners. A mix of long-slit and MOS programs plus standard
star observations. These will also let us shakedown all the observing procedures, scripts, faint and bright targets, and go deep enough
to get good numbers on backgrounds for exposure planning.
For these data tonight, we will use PARNTER=LBTSDT, since that is the closest category we have for the archive that will allow all
the partners to see the data. Individual programs have SV_xyz PROPIDs and each has a PI_NAME set in the scripts.
All times are UTC
unless otherwise noted.
- Before sunset, doors open and mirrors trying to cool to ambient. Sunset in about 30 minutes. Winds 8m/s SSW, 6°C, humidity 18%,
pressure slowly rising. Mostly clear, a little hazy out. Transparency OK for spectrophotometry.
The plan is to take some twilight spectra, particularly a slitless dual-grating as a pixel-to-pixel flat field method. Will also get twilight spectra
with the same slit configuration to compare with the twilights taken during the April run.
- Taking slitless sky spectra with the gratings in dual mode. Red is mostly a wash, but blue is terrific for doing pixel flats for spectroscopy.
We took spectra with the ND1.5 and Clear filters. All had ~30K-40Kadu, so we were guessing pretty well how to track twilight.
- Ready to go to a pointing star. GCS did not see the star. Turned out because Geno had done a closed-dome preset earlier while MODS was still in calibration mode, the AGw stage faulted (the AGw stage is interlocked out when the calibration system is deployed in the beam), and GCS did not recover/reset correctly. Once I reset the AGw stage and drove it to field center, it was fine thereafter
- First collimation star, BS9153. Mirror temp is 0.2°C different from ambient, so we're good to go.
First images were way out of whack as expected, lots of Z4 and Coma on the first corrections. Converging well, but it the GCS guide image looked funny, like it was subtracting background?
Midway through the converging WF, the star vanished. We looked at the raw images, and the frame is being heavily over-subtracted. Luckily Norm Cushing was in the house and we asked him if the experimental guide camera bias/flat correction bits were enabled, and how to disabel those. Many images were effectively filled with zeros, basically negative pixels on subtraction clipped to 0.0.
We were aware that data were being gathered to investigate if bias or flat images for the guider could be used to mitigate some scattered light artifacts during bright moonlight, but we had no idea it had advanced to implementation in the BP11 GCS, much less released for operations! We were never consulted on that implementation, nor informed that it was in the system we were using. That it completely failed out of the box makes matters worse. We should have been told, and it should never had been turned loose in an operational system like this without testing, evaluation, and discussion of the results of those tests.
- Torsten was called, and he turned off the image processing. We resent the preset and now the system works as it did before.
Seeing is very good, 0.6-0.8 arcsec.
- First science target: GD140 spectrophotometric standard star
Some oddities - the WF image was way out, but we just collimated. On doing an offset to center the star in the slit, the first image after the offset was terrible, with huge WFE (800nm vs 80nm before the offset), but it converged quickly. We only moved 8-arcsec in E/W so we're perplexed. All this is causing us to be slow off the block so we're falling behind our program.
SDSS J1141 QSO Pair
- Acquiring Hennawi's SDSS QSO pair, targets are to be aligned together along the 0.8-arcsec long slit. Seeing is about 0.8 arcsec.
Again, on offset to align, the guide image was awful and steadily converged (again, ~800nm WFE compared to ~100nm before the offset). Some minor bugs in modsAlign were rooted out (accidentally sent a double offset, oops), and we got beautiful alignment along the slit of the two QSOs with 0.8-arcsec seeing. A two-in-one slit acquisition is challenging, and it went very well.
- slowed down by bugs, but we got 3x20min integrations started. One source of the slows was waiting for WF convergence due to the problem of blow-up after offset.
Sadly, seeing started to degrade through the the last two 20-minute integrations, as well as getting to higher elevation. Spectra clearly show the Ly-a emission line in the higher-Z of the pair, the other has Ly-a redshifted right into the dichroic cross-over, which is why we'll follow this with a short red-only spectrum to cover that region.
We also noticed a lot of problems with AO convergence. Andrew Rakich called up to note that the PSF Arbitrator was misconfigured for Primary instead of Split mode. This could explain a lot (I won't try to reproduce Andrew's analysis). After the last spectrum is done, Geno will clear AO and correctly configure the PSF arbitrator for Split mode.
- AO cleared and PSF arbitrator reconfigured. Resending the preset and starting the red-only 2x300s integrations.
This time the offset went off without a hitch. The problem was the mis-configured arbitrator. Split mode is needed for MODS!
- Started 2x300s exposures. Getting to low elevation, so seeing is degrading
- Bootes LBG MOS field (Bian). A couple of minor glitches getting started, but 3x1200s integrations started. SEeing
is much better, 0.5-0.7 arcsec and good WF convergence. Much more stable than before. Can faintly see some of the targets in the first
- BD+33 2642 - Spectrophotometric Standard Star
1.01 airmasses, dual grating and red grating. Acquisition sufficiently well-centered in the 5-arcsec slit we did not refine the pointing.
Seeing is now 0.6-arcsec FWHM.
- Q1537 (OSU CIV program). Preset to PA=0, but this is an add-in because of the early delays, and we had not yet assigned this
a parallactic PA, so we had to abort the acq sequence after the preset, edited the acq script to put in PA=20° and a new guide star,
and then resent. Using the LS5x60x0.8 slit. Seeing is ~0.8-1 arcsec, some puffiness unlike for the past hour...
Seeing started out good, but degrading a little and the background is coming up a little (haze?).
- GCS Guider failed with timeout waiting for guide image. AzCamServer
was looping with something about a stuck counter. Followed
the procedures to stop GCS, reboot azcam, restart GCS to recover.
Final 1/3 or so of the last image was guiding open loop. Unsure of the impact.
- IGRJ18485 field (Masetti) - long-slit. Aligned to PA=0 for parallactic (we're close to transit). Finder isn't very good, so we are doing our best guess. There are two faint stars at the location of the finder, luckily for the PI both are aligned along PA=0 and separated by about 1.2-arcsec. Started 2x800s integrations. Seeing is 0.6-0.8 arcsec, so we're using the 0.8 arcsec slit. Some high clouds starting to come in on the satellite.
- MG4 200112+4352 - VHT BL Lac (Heidt).
We had a 1h error in the guide star RA, and the IIF didn't flag it until after
the preset had been executed. The guide probe position in XY coordinates can be computed anytime given the target and gstar coordinates, so why not do that and check for valid GProbe position and then execute the preset (or not) would save a lot of time?
Seeing is 0.8-1.0 arcsec, using the 0.8-arcsec slit at the PI's request. Oriented along parallactic 45° Object very easy to acquire, and gives good signal on the spectrum. This is good, as thin clouds are encroaching.
Note that the guide star is in the science field and partially vignettes the bottom of the long slit. Not a problem as we have lots of sky available to extract.
BD +28° 4211 Std Star
- Last target as astronomical twilight begins. Seeing ~1 arcsec, using the LS60x5 spectrophotometry slit.
Definitely clouds around, lost our guide star at the end of the first exposure.
- Ending 8Kx3K bias frames while Geno is closing up shop.
- 30 May 2011