MODS1 Commissioning Run 6: UTC 2011 March 18
Observer(s): R. Pogge (OSU), P. Garnavich (ND), M. Pedani (LBTO)
Telescope Operator: S. Allanson (LBTO)
AO Support: A. Rakich et al. Remote from Tucson
Telescope Support: J. Urban (LBTO)
Instrument Support: D. Officer (LBTO)
Worked most of the night except for 3 hours in the middle of the night lost to high winds. Got guiding stability and accuracy tests done, successfully aligned an MOS mask with the prototype modsalign script, and did a couple of oddball long-slit acquisitions with spectra to walk through the sequence.
MODS is working well - the active optics problems of previous runs now seem to have been solved - we got consistently nice, round images even with reasonably off-axis guide/WFS stars. Still some details to work out over the next couple of nights if the weather and telescope behave.
Conditions were windy, with bright moonlight (1 day to full) and thin wisps of cirrus but no heavy obscuring clouds. Seeing was very variable, between 0.5 and 3.5 arcsec, but most of the time we were 1-arcsec or a little better.
Dewars filled at 3MST - still showing good holding characteristics (30h for red, 32h for blue).
All times are UTC
unless otherwise noted.
- Scattered high clouds and a few small puffies at mid-level. Warm with temps near 8°C and humidity 25%. The main weather concern for tonight will be wind: all afternoon we've recorded SW winds 10-15m/s with occasional gusts to 18m/s.
- Sunset, still windy but trending downward. Thin clouds off to the NNW, and blowing over. We're open with tail to the wind to try to equilibrate some. Not sure what seeing will be like yet, and whether the wind will behave itself and stay below limits.
- First pointing star for the night.
IE = -60
CA = +85
Collimation star towards W to avoid facing into the wind, using BS9122 (dec +44).-
Converged pretty quickly to 150nm rms WF error, seeing 0.7 arcsec. Not bad for a start. The temperature is not different than last night, so the mirrors doesn't have far to go to equilibrate. Seeing is variable, 0.7-1.0 arcsec w/occasionall puffies.
Guiding Stability and Accuracy Test
- LST is 0635, going to a faintish blue star, CSS081231, an eclipsling CV, it'll be a good target for guiding accuracy. We'll center it up in the 5-arcsec fat slit and track on it active preset for an hour or so, taking images in ugriz
Guide star is about 2.5 arcminutes off axis.
- seeing is good, 0.6-0.8, so doing a slitscan first, then we'll do a long tracking run. Images mods1r.20110718.0007-12
Nice round images, 0.6-arcsec FWHM at SDSS i
Wind is now under 15m/s, but we're watching it all the time.
- first trackTest.obs sequence: 15s r+g, 30s u+i, 30s u+z. Star centered on the LS60x5 slit. IMCS in continuous operation (all exptimes are >10sec). IMCS lock-on requested at the start of each script, then engaged during exposures. We can use the edges of the LS60x5 slit to track IMCS performance over the sequence, and measure star drift relative to the slit over a 2 hours.
images: mods1b.xxx.0004-6 mods1r.xxx.0013-15 - nice round images, 0.7-arcsec seeing
Starting at 0247 running trackTest.obs inside a shell script that will execute it 10 times in sequence. seeing is variable. This will give us a roughly 45-minute tracking sequence. So far we're seeing about 5 minutes from start-to-start for the sequence, which is about right.
At 0300 the seeing started degrading to 1-arcsec FWHM but still tracking. Not seeing WF errors change much, so not (we think) image degradation due to active optics. This will be a good test for Andrew and Doug to look at since this is a 2.5-arcmin off-axis star, and not aligned with the Y axis.
Started next 10 sequence at 0330 UTC, this will give us ~90+ minutes of tracking from roughly meridian down. We started at elevation 79° (Airmass 1.02) about 7 minutes before transit. Guider reporting seeing 1.2 arcsec FWHM, but not seeing any chagnes in RMS WF error. Images are still very round, just bigger...
- seeing just puffed up to around 2-arcsec, definitely degrading from the nice seeing we started with... Wind is starting to pick up again, could be correlated. The bad seeing puff was around the time the wind gusted briefly to 18m/s, but we've seen worse seeing uncorrelated with wind. Oh well.
- Temperature outside dropping, so we could be getting some dome seeing (delta-T around 2°C inside to out). Differential with mirror and enclosure is only 0.5°C. Seeing now consistently about 2-arcsec FWHM. RMS WF error still around 150nm.
- Images are still nice and round, however. Only a brief period around 0400UTC when the RMS WF errors went up to about 250nm. Active optics is definitely behaving itself even if the seeing is not.
- glitch in comm from the red channel CCD control computer - it did not send the "exposure" done message, even though it read out the CCD and saved the image to disk. Noted the time to check engineering logs.
By the way, this star went into eclipse in the in the past couple of minutes. Eclipses last 5-7 minutes every 117minutes. We chose the target because it was very blue, in the right Dec range, and we had a finder. With no comparison stars inside the LS60x5 slit we won't get photometry, but it makes watching the otherwise boring tracking sequence more interesting. Eclipse depth is ~5 mag in the u band (total eclipse of the white dwarf including the accretion disk hotspot).
Low elevation differential refraction tracking test
- Done, going to BS9107, in the NW and setting at about 28° elevation. This is for the differential guiding test. First doing an on-axis active preset to set the collimation, then selecting an off-axis star and putting in the LS60x5 slit. Seeing is 2-arcsec (or worse). WFS convergence is a little slow, but getting there.
- Offaxis star similar to the last pointing (2.5-arcmin and SE at roughly PA30). Seeing 2-3 arcsec according to the guider, taking g and r snaps of 10s exposure to check centering in the LS60x5 fat slit. PA=0 (we could orient along parallactic, but that's hardly sporting).
Applying a DETXY offset of 0.54-arcsec in X, none in Y. Blind pointing is pretty good! We're using the SDSS r filter for centering on the slit because that and g bracket the guiding wavelength of 565nm.
- starting, 30s in each filter, starting at elevation 25° (airmass 2.3).
Starting a 10 script sequence at 0445. Note atmospheric dispersion is such that at SDSS u the star is nearly off the slit. Most of what we want to measure will be from the riz filters.
Seeing is awful - 2.5-arcsec FWHM, up to 3-arcsec at times. Amazingly the GCS still manages to find something to lock on most times. WF errors are around 200nm RMS or less.
Just noticed that the AGw guide camera has the F525LP
filter in, so the guide wavelength is really more like between r and i. Interesting...
Stopped the sequence at 0512 UTC, when we were at elevation 19° (airmass 3.07).
- On to BS9139, high elevation, to collimate and see what the seeing is like at a more rational elevation (65°).
We'll also get a better idea of the true wind speed as we've had the anemometer in the building lee for a while, and are hearing rumbles from gusts.
The star was pretty distorted when we went to it. Time about 0517UTC, might be interesting to Andrew and Doug? We're letting the WFS have at it. Could at least be entertaining.
Steven had to clear the active optics correction as it was showing signs of diverging. Then it started with collimation model and sent to 0.8-arcsec FWHM and reasonably round! See Steven's notes for details on what he did.
Wind is picking up, we have had a couple of gusts up over 18m/s in the past 10 minutes.
Pointing Stability Test
- Marco is running a pointing stability test; 5 stars from NW to SE. We're orienting the telescope with its back to the wind. We've had a few gusts above 20m/s
but sustained is still below
Pointing test using the faint star catalog. Each slew is a track preset followed
by a 1s snapshot with the MODS red science camera, SDSS r filter and 1Kx1K ROI, performing
a manual IMCS lock before each image.
0543 - Slew to WT10_348F to check Pointing: MODS1r Image 0098
0546 - Slew to WT10_383F to check Pointing: MODS1r Image 0099
0549 - Slew to WT10_352F to check Pointing: MODS1r Image 0100
0553 - Slew to WT10_309F to check Pointing: MODS1r Image 0101
0556 - Slew to WT10_275F to check Pointing: MODS1r Image 0102
Pointing RMS Scatter Results:
sigma(CA) = 3.4-arcsec
sigma(IE) = 2.6-arcsec
- Wind starting to gust around the limits. We're orienting the building to best monitor the true windspeed, and doing an active on-axis preset to collimate and settle while we watch the weather (the pointing test used track presets)
Seeing is about 0.8-1arcsec FWHM, but WFS error is ~400nm up from about 180nm.
- A different star now, better place to measure the wind...
- Wind hasn't been definitive yet, so we're going to the Abell 1689 MOS Test field in the SE. This has the wind somewhat flanking us, but should be OK as long as the scope doesn't complain
- Preset failed, no star. Steve checking the pointing.
IE = -40
CA = +91
- Trying again - and failed again. This is the same .acq file as last night. Strange.
- going after a definitively bright M5 pointing star, maybe we grabbed the wrong star?
Not there, it was a bit off...
IE = -82
CA = +87
- and again... much better.
- Just starting to converge on the WFS when the wind gusted above the 22m/s limit. Steven is closing up
Enclosure light tightness and internal background test
- Since we're closed, we'll do an instrument light-tightness and internal background test. Winds now sustained 20m/s, so we have some time.
Step 1: Enclosure lights off in a closed enclosure at night.
- Instrument in dual imaging mode
- Full-Frame readout and Clear filter
- Hatch closed, imaging mask, and calibration tower and AGw stage retracted
- Set of 5x300s open-shutter integrations inside an ostensibly dark instrument
- Instrument at rotator angle 207&\deg; (nominal PA=0° @ meridian)
Step 2: Turn lower chamber flood lights on (not the bright overheads) and allow about 10 minutes for them to warm up and get to operating
- Same configuration, repeat sequence.
Median the 5 frames together for on and off, and form the difference (on minus off) for the median on and the individual on frames.
Examination of the data shows a very small light leak in the red camera's view of 0.012 ADU/sec, with a peak of 0.020 ADU/sec in one coherent area with lower chamber floods turned on.
The blue camera shows a mean level of 0.052 ADU/sec, but a region of an apparent leak at the right-side of the CCD of nearly 1 ADU/sec. This may
be looking at a leak through the enclosure - most likely at the instrument/rotator interface - on the red-side of the top of the instrument. In the instrument orientation during this test (210°), this is the side of the instrument that is directly illuminated by the lower chamber floods. The region of the ball/socket joints at the instrument/rotator interface are known to have a few leaks We will investigate this further this week.
Typical flat-field exposure times are of order 1-10 seconds, so total counts from leakage from the lower chamber floods would be under 1ADU in red,
and of order 1-10ADU in blue at the far right (red) end on full-frame spectroscopic images. Compare this to desired levels in flats of a few thousands of ADU.
In principle one could take flat fields with the lower chamber floods on. We have taken closed-shutter bias frames with full chamber lights on
(including bright overhead halogen lamps) with no measurable parasitic light component.
Internal background test analysis uses the lights-off frames. We'll do that later since the wind is changing in our favor.
- slight downward trend in wind speed for the past half hour. Still not able to open, but maybe later?
- the wind has been below threshold for nearly an hour, so we are setting up to open again. Unfortunately our primary MOS test targets are now moving into the SW into the wind, so we shall see how well we do.
- Active preset to BS9152 to collimate on-axis
image badly aberrated. Steven clearing active optics. Looks better, starting to converge, putting in a bunch of focus and coma corrections at first
Mirror is 0.4°C different than chamber ambient, but nearly 2°C above outside ambient. We may have problems with dome seeing. So far the WFS has not been very stable, seeming to converge then back up to 500nm rms. Seeing is variable, sometimes down to 0.7, others up to 1.2-arcsec which does not help.
MOS Target Acquisition and Mask Alignment Testing
- Stable, going to the Abell 1689 MOS Test Field. This is into the wind, but the wind speed is now down around 10-12m/s.
- Shooting a 60s thru-mask image, SDSS r and red camera, and a field image (mods1r.xxx.0123 and 0124)
Analyzing with modsalign.py, so far so good, no problems (no repeat of last night's fun) Using 6 stars for this fit.
offsetpointing 0.0515 0.27 -0.89 detxy rel left
to the telescope, and once settled taking a confirmatory shot: mods1r.xxx.0125. Seeing ~1-arcsec but variable.
Looks good (right rotation sign? hard to tell the angle is so small). We are going to now try to take spectra and see how well it did.
Doing everything by-hand with inspection, so a lot slower than we hope to be in the future.
- firing off 2x600s spectra.
Seeing ~1-arcsec, WFS stable and about 150nm RMS. Guiding and exposing...
- Following with a 600s dark-mask spectrum to get the moonlight background component.
Moonlight didn't subtract as well this time, it is more time-variable. We really need that baffle...
- Back to the A1689 mask - how repeatable is it on just mask exchange. Doing just one 600s exposure this time so we can compare 1:1
with the previous ones.
- Thru-mask image after the spectrum following the mask exchange - 60s and 120s
Long-Slit Extended Target Acqusition
- Preset to M101. It is a target to do an extended source long-slit acquisition to see
how well it goes compared to a star acquisition (which is what we've done so far). It is also in a good position with our back to the wind. And it is a big, cool face-on spiral galaxy, so the spectra might be interesting to look at.
- Started spectra 300s. Took 15m from point to start, a lot of stuff done by hand and so inefficient (plus a couple of ways of estimating the offset to apply).
Wind is to our flank, still SW at 8-12m/s, so calming some from before.
Taking a 600s exposure. Moonlight still bad, but the moon is setting so it is getting better than with A1689.
- Post spectrum thru-mask image to measure any jitter in the past 24 minutes.
Target Acquisition without a finder, into 0.6-arcsec slit
- always fun. We're going to see if we can find this supernova target and get it on the slit without a finder, trusting coordinates and DPOSS.
Had to unwrap in azimuth, taking longer than usual to preset... Seeing ~0.8 arcsec before start of WFS correction
First snap at 12:19, started spectra at 12:28, again working by hand. Offset was small, about 0.6-arcsec in X, none needed in Y. Target was actually pretty obvious, so it wasn't very sporting.
- Getting light out, we're done.
- 17 Mar 2011