TMS Meeting Minutes, March 25, 2021

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

Last Week's LBT Visit (Heejoo):
  • Fibers and collimator install
    • Fibers routed on the SX side
    • Collimators almost fully aligned, except for 26 and 27 (and two others)
    • No time for laser tracker survey
    • The pointing needs to be adjusted for the aligned collimators
    • Leroy or Steve Ellison would both be capable of working on the alignment
    • Heejoo asking Leroy to install a retro on fiducial 25
    • Currently, we are using the F280FC-1550nm collimator for all spots on the SX side
      • This is a monochromatic collimator... the 633 nm spot is massive
      • This collimator focuses the red light at about 600 mm
      • By implementing a long radius lens to counteract the focussing, a drop in could be used to maximize red light visiblity at 10 m
  • Wide angle RR test at 2.4 m and 10 m distances
    • Compared with off the shelf retro from Newport
    • The wide angle RR gives 12% return at normal incidence, and 10% return at 50 deg incidence (while there is no return at 50 deg from the Newport collimator)
    • These are a measure of the contrast at the fiber tip
    • Though the overall intensity is less (max 14% from RR), the contrast is higher because of intensity matching between RR and fiber tip
    • Some dropoff for 10 m for both wide angle and the off the shelf RR's
    • It is clear that the wide angle RR works for Etalon, allowing for edge sensing and TMS operation
  • SNOW!
  • Took twice as long as usual to ascend the mountain
  • Remote operation of the laser tracker metrology system is quite feasible (Heejoo and Andrew have confirmed this)
    • Using hot glue spots on the feet of the tracker takes the least-squares tracker repeatability error from around 500 microns to around 8 microns

  • The updated Etalon software appears to be working well with no issues

  • Order in for collimators from Thorlabs ($320 per)
  • Lead time is about 2 weeks
  • Planning to bring the new collimators to the summit to test during the next visit

  • Andrew has a new wide angle RR
    • Doesn't use a beamsplitter
    • Three glass types
    • Monocentric (with spheres), symmetry (Schmidt principle)
    • ~160 deg full angle

On going from one-sided to two-sided operation:
  • Yang assumes you could just launch the sides separately
  • In an earlier version of the prototype, going from one to two sides wasn't terribly difficult, according to Andrew
  • John remarks that in the best case, you want to make one set of measurements and compute corrections afterwards
  • SX side should be operational in the middle of April (pending Heejoo's next visit and the laser tracker survey)
  • Yang plans to tweak the current version to make it work for two sides

March 19th Data Analysis
  • Heejoo looked into which channel experienced threshold issues
  • Counted how many times the nine channels dropped due to threshold constraints
  • Most channels were around the same, but Ch. 23 was higher than the others (about ~60 extra events on Ch. 23)
  • Some question on the method of counting the threshold dropping flag
  • Andrew recalls that the system was fairly stable on this particular night (TMS was run passively)

Three more hours of active time on upcoming Sunday night (9:50 MST to 00:50 MST, 4:50 to 7:50 UT)
  • Andrew would really like to run VFast Extra while controlling actively with TMS obtaining quantitative data from FPIA pupils
  • Select suitable fields optimized for FPIA operability with nice target stars
  • Planning to run the loop to obtain pupil images and post analyze
  • Red side gets active control from TMS, blue side runs open with no active correction after initial collimation with FPIA
  • Goal is to see if TMS can successfully keep the system from defocusing over time

Yang Comment on Software
  • Yang thinks we should talk to Etalon and convince them to redesign their software
  • As of right now, their software is unable to be operated remotely
  • In the future, this will pose many problems for operation of TMS
  • Yang thinks they should separate the TCP/IP from the Etalon GUI with a lightweight program for listening to open port
  • This should be fairly simple to implement, where all control flows through TCP/IP without relying on the software
  • As of right now, you can only control the major operational parameters of the EAMT from within the Multiline software, which is very limiting
  • This sort of operation could be VERY USEFUL for other Etalon customers
  • In Andrew's experience, Etalon are very reasonable people and willing to consider new ideas
  • Andrew proposes writing an introductory email with Yang to discuss the problems of interest
  • Heejoo also wants to discuss shutter operation, pilot laser operation, and other topics with Etalon
  • All of these things should be accessible outside of the Multiline software
  • Yang proposes we start this early and get on top of it right away, as this could be very problematic for future operation
  • Frame this as "LBT is seriously considering implementing this long term into the observatory systems, and it would be really great if we could have the needed functionality"
  • Etalon may also suggest an alternative solution, which is fine, so long as we have the necessary operation

Recap from last meeting
  • Yang starting work on transitioning TMS software output to a database
  • Will be implemented into the DMS web system
  • Could even reuse some existing tools to allow for browsing and exporting of data
  • Next step, developing an integrated GUI version that's as simple as possible
  • Yang proposes that TMS is two primary operations
    • Start TMS by taking a reference then running
    • Start TMS using a previously capture reference
  • For running TMS with a previous mirror reference positions, Yang suggests we default to using the last mirror position, rather than having the OSA's choose (start with current, or start with last known)
  • Also planning to include a pause button, which keeps progress in memory, and then continues on with saved reference
  • Advanced options for advanced users: toggle TMS on/off, reset the reference, drop channels, adjust thresholds, etc.
  • The GUI is also a client, so a command line client will be produced
  • GUI and command line client will share the same code, just operate with different controls
  • Some coordination needed, as you don't want multiple TMS processes running simultaneously
  • Only one copy able to run at a time; if a GUI is running, can't use command line
  • John requests making this like the TCS subsystem, where you can have as many GUI's as you want, all talking to the same system
  • Could also be like ECS, where you have a "read only" option, where you can open the GUI and observe without endangering the process
  • Rely on human coordination to ensure there are no conflicts of interest with multiple corrections
  • Yang suggests having some regulation to help manage read/write operations of the GUI
  • John would really like it to work under the TCS model, where either a GUI or a script can send commands to the subsystem, and the subsystem sends commands as they are received
  • Yang concedes that multiple users can interface the GUI in their own instance, with the ability to control from multiple points
  • Yang planning to prototype an interface, receive feedback from TMS team, and then implement it

Regarding system readiness
  • Andrew believes that we could present a system that is ready for science by May
  • John hopes this is true, but he reminds us that we have said this multiple times, and consistently found some new problems to deal with
  • Regarding the focus correction issues, there were some issues with TMS defocussing with different truss configurations, which was proven to problematic
    • This was based on some poor assumptions about noise and the inconsistencies when switching from 9 ch to 8 ch operation
  • However, there is still a standing question about the slow drift of focus that LBC red showed during the 20201201 dataset
  • Olga's remaining concern is about a buildup of spherical when focus is degrading
  • John believes that the script is correct here, that there isn't an issue
  • One of these days, it will all work, and we'll hold the focus steady for five hours
  • As a reminder, this is a whole new way of controlling a telescope
  • Controlling to a micron at a distance of 10 meters (absolutely bonkers)

-- TrentonBrendel - 25 Mar 2021
Topic revision: r1 - 26 Mar 2021, 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