SX On Sky Checkout

  • The clouds are starting to build up and the seeing is marginal, so they are doing the PEPSI FPU stuff.

    We finally achieved a successful bin 1 preset with the new 400 mode reconstructor. I had to restart processes a couple of times on both machines and ended up killing a bunch of old processes running under AOeng, as well as all ssh clients. That seemed to do the trick.

    The loop was skipping when I applied NCPA. I turned NCPA off and re optimized gains. Everything was running well, but then the clouds moved in... Seeing was probably the cause of the NCPA problem (The NCPA amplitudes being applied looked correct and the GOPT process was working well). It was at and slightly above the 1.2" limit and I think the pupils were losing too much light. This will need to be verified, of course, but the weather is not cooperating.
  • Things are running smoothly in bin 3 except during some offsets. When the loop resumes, the adsec has a funny shape on it, causing a double image on the GCS guider. In the CCD 39, it appears that there is a dark, nearly horizontal line bisecting the pupil images. When the telescope does a subsequent offset, the loop resumes properly.

    I've seen something similar when observing with Jochen, this spring. Some offsets would appear to start before the loop was fully paused, causing skip frames as the loop would try to compensate. The problem obviously isn't that bad this time, but I wonder if it has a similar cause. By the way, the above is more of a hunch than anything observed.

    I took some optical loop data during one event: 20130913_064837
  • 20170913_065505
  • During one of these "events" the loop opened with the safe skip. Even when resetting the default, 653 mode flat, the dark line was still there. I'm now confused.
  • Another dataset 070126
  • New Target: Bin 2. Optimize gain over-drove the mirror. The same thing happened in Bin 3.
  • The edge looks like it wants to roll. Dataset: 072118.
  • Disregard the edge roll comment. Those pixels are outside of the pupil. Indeed this binning is working fantastically.
  • Good manual gains: 2.00 1.10 0.65
  • Below is the rest of my log, I'll put it into the WIKI later (It might be cloudy the first half of tomorrow night). Things did seem to run better after I changed that IP address to the correct one. (192... instead of 193...). I didn't see the same issue tonight as last night, so maybe I'll set that to monitor. Anyway, I had to look through the issuetracks to remind myself of how to recover CCD39 (there were too many instances of the program running). Other than that, the AutoGains were immediately overdriving the mirror, in every binning. I don't know if anybody will have time to look at that tomorrow (today). Dave T is taking over, so we're finished for tonight.
  • 073204, 073343, 073546, 074224, 075612
  • Loop stuck in pause, loading new script to finish object. Had to do it as a new preset. The offset seemed to miss the star. I could see it in one pupil image, but only partially illuminated.
  • Bin 1 this time:
  • 081314, 081335, 081415, 081442
  • 081637, 081704, 081723, 081741
  • Seeing bubble, reclosed loop, twice.
  • Gains too high. Autogains not working, I set them manually, but they were too high. Fixed 'em.
  • Gopt 0.3, when I did the "renormalize"
  • Gains: 0.41 0.22 0.06
  • 084145, 084758, 084813, 084833
  • Safeskipped right before an offset.
  • The Secondary was at a limit.
  • I found out something new about "Intervention Mode". If you don't click on any of the buttons in the last set except "Done", you can't re-close the loop.
  • Going to Al's Asteroid
  • Mag 8.1
  • Same gains.
  • 093320, 093338, 093356 loop might have been open, 093421, 093442, 093505
  • Going to a bin 4.
  • Manual gains 3.4 1.8 1.3
  • 095715 Too many frames, it spanned several offsets and an error (non failure) 40000 frames! oops,
  • 100409, 100432, 100451, 100508
  • We occasionally came close to the off-axis limit of the AGW. The script continued, but it stayed paused during that position.
  • Back to the Asteroid
  • Same gains as before, smooth as glass.
  • An issue or two occurred, with finding the asteroid again. It is working now.
  • Doing a standard star.
  • bin 2 400Hz
  • 111816, 111837, 111903, 111949
  • It's a double star!!!!! Unlisted anywhere that Barry can see.
  • Back to asteroid, smooth sailing.
-- %USERSIG{GregoryTaylor - 2017-09-14}%
Topic revision: r1 - 14 Sep 2017, GregoryTaylor
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback