TMS Meeting Minutes, February 10, 2022

Attendees: Andrew, Breann, Christian, Heejoo, John, Olga, Matthieu, Trenton, Yang (Zoom Meeting)

Updates:
  • One night for passive mode data acquisition
  • Very poor seeing, took lots of LBC data out of focus
  • TMS measured pose during long measurement period (no star images, just out of focus images)
  • The Etalon team asked for 24 hours of software testing starting tonight, ending tomorrow
  • All fibers are unplugged so as not to affect observation
  • The fibers will remain unplugged until Sunday morning, when LeRoy will complete realignment of the collimators
  • LeRoy tried to realign the collimators last weekend, but didn't have enough time
  • LeRoy found out that one of the screws on Ch. 5 collimator mount was loose and the collimator was floating freely
  • Christian stated that there is a possibility of conducting TMS testing by 3 PM if the alignment is finished
  • There will be backup plan in case things don't work out, but we'd love to use the time on Sunday afternoon for TMS testing
  • It may be best to wait a bit until the beginning of the night when the temperature is varying greatly
  • There is some question as to whether it is worth running the TMS towards the second half of the night
    • As it stands, TMS is the backup in the case of worst conditions
    • The cases where we have large front and back temperature differences are the most interesting to Andrew (possibly at the start of the night, but it doesn't have to be)
  • Heejoo will keep an eye on the aligning process and will distribute an email around the proposed TMS handover time to update on status
  • Heejoo notes that the TMS collimator alignment usually takes 3-4 hours, but there aren't other plans on the telescope on Sunday morning
  • Heejoo hopeful that we will have a fully aligned TMS system by 3-4 PM on Sunday
  • Heejoo will be available on Sunday as we attempt to test TMS
  • Andrew recommends considering using VFAST Extra if seeing is really bad

Detailed Data from 20200127 Passive Mode:
  • Passive mode data collection on January 27th
  • Very noisy data noted, so Heejoo analyzed the std. dev. filtering and which channels were dropped
  • It turns out that there are lots of std. dev. data errors
  • It was proposed to look at the std. dev. data only during exposures
  • During exposures, we most likely see stable relative position between M1 and LBCr
  • There were some points far exceeding five standard deviations, which Heejoo plans to look into both in pose and in image degradation/elongation
  • There is some possibility that these jumps in the std. dev. could be LBC images with filter changes (there is a Z-offset added with a filter change)
  • John encourages Heejoo to look into the filter configuration as well to get to the bottom of what these std. dev. jumps might be caused by

20220207 Data:
  • Huge changes (~35 arcsec) in Rx around 4 UT, with no change in Y, on Dx
  • John notes the same thing on his LBT primary collimation plots for Dx
  • Olga notes that there was a large manual correction for coma, Z7, and focus, Z4 (just Z4 on left side, Z7 and Z4 on the right side)
  • At 04:09 UT, added Z7 correction of 1500 nm
  • At 04:11 UT, added Z7 correction of 3000 nm GLOBAL (+1500 nm)
  • These corrections are global Zernike corrections (global shape corrections)
  • Through the PSF GUI, when you correct coma and focus, you would expect the mirror to move both in Rx and Y, not just Rx
  • Should be a rotation about the vertex center of curvature of M1, Andrew notes
  • Andrew also notes that the coma neutral rotation points for the Gregorian and prime focus modes are different
  • The pointing neutral rotation for the hub is a rotation about the vertex of the primary mirror, as it's a first order system
  • However, with a manual correction, you could have a bend applied to the mirror
  • Heejoo curious about how well focus was kept over time, and if it's possible to have such a large change in dRx with no change in dY
  • Olga notes that there is a factor of 21 arcsec relating Rx, Ry, and Rz to X, Y, Z, which is included in Heejoo's plots
  • Suggests data processing over time to extract FPIA data offline from defocused images
  • Olga can only start this task of offline FPIA analysis after 4:30, because before that there were in-focus images
  • Seeing was so bad that they couldn't just collimate and lower the mirror by 0.8 mm and expect to have an evenly illuminated pupil image
  • Jump at 5:20 from Z4 introduced
  • Heejoo planning to look into the continuity of the PSF between the two distinct regions of data

20220209:
  • DIMM seeing was great, beautiful weather, and LBC did superb job
  • No TMS, as fibers were unplugged and collimators need alignment
  • Mirrors were struggling to stay in thermal equilibrium, especially in the early morning after the termperature rise from 10 UT to 11 UT
  • Lots of higher order spherical being driven in by thermal gradients
  • FPIA mistook the bright core produced by high order spherical as the edge of the disc; this required manual correction to ensure proper collimation
  • The last field actually had very good quality; Christian plans to update the IQ plots to incorporate these images, as they were disincluded before

Smoothing Out Performance of TMS:
  • When we are down to the point of making small corrections, gain down to 0.05
  • Andrew proposes that, once we reach the ringing cases, we could apply a damping factor of 0.5 to damp out ringing
  • If our radial correction from TMS is less than 10 um on an R = sqrt(x2+y2+z2) basis, we could damp out the noise
  • Andrew notes that this TMS ringing could actually degrade good seeing and it should be fairly easy to damp out
  • John believes that it, in the long run, it's a good idea, but he is not sure if we need to add that complexity right now
  • Yang thinks that this should be fairly straightforward, but requests that Andrew document what he thinks needs to be done for Yang's reference

20220101:
  • No elongation on blue side, only elongation on red
  • Andrew wonders if there was a period of time with no ocrrection and then a large correction after a pause
  • If TMS perturbs the telescope during exposure, we haven't achieved our goal
  • If TMS is able to recover from an external perturbation during exposure, TMS is achieving its goal
  • It is quite difficult to force TMS to stop corrections during an exposure, according to John
  • If TMS corrects a drift that's occuring during an exposure, that's great
  • If, however, the drift is allowed to progress over an exposure and is then corrected partway through by TMS, that may be less desirable
  • If we know we have been dropping measurements, we may need to consider how TMS handles this gracefully
  • Elevation changes always causes huge std. dev., while thermal drift causes changes that are an order of magnitude smaller, but could still cause issues with the 5 um std. dev. filter
  • 06:44:47-06:48:42 dropped many channels all at once, and no corrections were sent
GMTO + LBTO TMS Colloquium in Tucson:
  • There may be an opportunity to have a colloquium for GMTO and LBTO to discuss TMS in early April
  • Andrew will keep us posted
  • Also some interest in going up the mountain, but John notes that this will only be allowed if there is work being done

-- TrentonBrendel - 10 Feb 2022

This topic: Software > WebHome > TMS > 20220210_TMS_Meeting
Topic revision: 11 Feb 2022, TrentonBrendel
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