TMS testing with binocular LBC science - 20180119 UT


Here's a short summary of the TMS-assisted LBC observations on 20180119 UT.

We started about 6 UT, and spend 1:20 hours sorting through bugs in John's python scripts, Multiline configuration problems, LBC issues, OB issues, OT issues and DOHYBRID/DOFPIA active optics. "We" includes Andrew C, Olga, Jenny, Shane, Steve, Andrew R and John H, and all these folks were needed to keep track of the observing and the flow of various data.

The good news:

Starting at 07:23 UT, we collected a time series of binocular LBC observations for the Kurtis Williams' TSIP program in M67 which lasted for 5 hours without any interuptions for LBCFPIA recollimation. Unfortunately, the cirrus clouds got thicker and thicker with time so the last 2 hours may not be all that useful for photometry of faint white dwarfs. There was more than 2 mags of extinction by 12 UT (at 2 airmasses). Seeing was 0.7-1.2 arcsec on the DIMM.

TMS worked fairly well for its first time ever doing a real science observation. Once set-up, it had only one hiccup where the software locked up and the PC had to be rebooted (without any disruption in collimation). We think that hang was caused by Andrew R paging back through some older data. There was also an instance we don't yet understand during setup where a bogus correction of -77 millimeters was received. With a TMS collimation update every 30 sec, we had a typical X/Y rms position of 5 microns. We think these residuals are caused by dome seeing.

For this type of time-series observation the system worked well enough that (weather permitting) we are going to try to run it again tonight (Saturday) using 3-ish people rather than 7.

Lots of wrinkles to sort out:

Unfortunately this system isn't yet perfect and we have lots of wrinkles to work out before it is ready for production science.

  • First, we are limited by how well DOFPIA can collimate LBC in the first place, so the LBC-Blue images started out looking like 1 arcsec kidney beans (not atypical). This is caused by the typical thermal "skirts" on the pupils that disrupt the FPIA pupil-fitting algorithm.

  • Next, TMS did not hold collimation perfectly for 5 hours, we had to pause TMS every 1.5 hours to correct by eye the drifts in focus (expected), spherical aberration (expected) and coma (not expected). Nonetheless, with these eyeball corrections we were able to maintain 1-arcsec-ish image quality the whole time with LBC-Red mostly better than LBC-Blue (common for LBC).

  • We expected the focus to drift with changing temperature as we have not yet implemented a focus correction based on the TMS-measured diameter of each primary mirror. A front went through about 10:30 UT dropping the temperature by 2 degC in 30 minutes.

  • We also expected the spherical aberration to drift as nothing is measuring it after the initial DOFPIA collimation. This will be a problem for programs at the beginning of the night.

  • Not expected was the drift of coma (up to 800 nm wavefront over 5 hours on LBC-Blue). Our first impression is that this is thermal distortion of the primary mirror, but alot of analysis is needed to demonstrate that as the cause.

  • There was also an expected drift on LBC pointing with time. The good news is that it was very smooth and steady. Beyond the normal open-loop tracking drift, TMS has to add a pointing correction that is linear with elevation to keep the primary mirrors within their travel range (relative to the normal LBC collimation model). Because the sky was so gunky at 12:30 UT, we were able to work into twilight to collect 17 stars worth of TMS/LBC pointing data.

Next steps:

We've collected a load of data by collimating with TMS every 30 seconds for 5 hours. That will have to be carefully analysed next week so we can provide more quantitative results than this email.

Andrew R also has some more work to do on beam stability and alignment by switching to a larger model of collimator.

Andrew and John need to work on implementing the focus correction algorithm based on mirror diameter, telescope temperature and dN/dT corrections by TMS.

More work on the various software packages is needed for better trapping of errors and bogus data.

Other possible improvements:

Should we average a few TMS exposures to reduce jitter from dome seeing?

Should we save error messages in the logfile?

Should we use averaged PSF temperature for the distance correction?


06:00 approximate handover from LUCI observing

sometime early: SX pointing vector was -0.789 -0.551 0.0 -11.8 +16.96 and DX was nearly identical. Why are they equal rather than opposite?

sometime later: SX pointing vector was +0.829 +0.532 0.0 11.42 -17.81 and DX was nearly identical.

07:23 TMS loop closed and time-series LBC observations begin

08:00 SX side is starting to show coma

08:24 SX +500 Z8 abs

08:28 SX +350 Z8 abs

08:30 SX -500 Z4

08:34 SX -1500 Z4

08:38 SX +1000 Z4

08:39:30 SX Z11 from +1000 to +600 abs

08:43 SX +1500 Z4

08:47 record reference and restart Blue TMS loop

09:58 Multiline software is stuck, will have to reboot the PC.

10:13 Back in operation. We think this was TMS operator error when Andrew was scrolling back through old data.

10:58 Add +1500 Z4 rel to each side

11:03 DX +200 Z7 abs

11:04 SX +100 Z7 abs, DX increase Z8 by 100 = 450 Z8 abs

11:07 DX reduce Z4 by 300 = +1200 Z4 abs

11:08 DX increase Z7 100 = +300 Z7 abs

11:09 Set new TMS reference

11:36 SX increase Z7 300 = 400 Z7 abs

11:44 SX increase Z7 400 = 800 Z7 abs

11:52 SX increase Z4 2000 = 5000 Z4 abs

11:55:30 SX increase Z4 1000 = 6000 Z4 abs

11:56 Set new left reference and restart both sides.

Pointing Stars

These pointing stars collected with TMS loop running the whole time.

12:42 ACT0254

12:xx ACT0290

12:49 WT10_292

12:52 WT10_339

12:xx WT10_326

13:07 ACT0332

13:xx ACT0356

13:13 WT10_328

13:15 WT10_326

13:22 WT10_294

13:23 WT10_243

13:26 ACT0190

13:xx WT10_105 TMS not settled?

13:xx WT10_105 repeat

13:20 ACT0177

13:34 ACT0230

13:36 WT10_281

13:40 ACT0311


Christian's image analysis is at:

-- %USERSIG{JohnHill - 2018-01-23}%
Topic revision: r2 - 24 Jan 2018, JohnHill
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