TMS Meeting Minutes, September 24, 2020

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


Meeting discussion summary:

Andrew's Results
- Mathematica results
- Took the nominal vectors and rounded to the nearest 100
- b1-b14 nominals all rounded to the nearest 100
- Used Yang's Mathematica code to test channel dropping
- Took 4 consecutive lines of data from eTalon output
- Nominal is a very different vector (formerly based on laser tracker measurements)
- Computes first pose acceleration (change in the change of pose) for nine channel change in pose between nominal and IP1
- Then computes for 8 channel, 7 ch., 6 ch., etc.
- When you do your first dropped channel, massive errors, like a reset, zero offset (like degrees and meters of error)
- Take this error and accept as constant for subsequent meas.
- If you use this as a constant correction, makes errors much smaller on continued dropped channels
- Basically, if we drop a channel, do the 8 ch. change vector, don't apply it immediately, but apply it to the next 8 ch. measurements
- In 7 sec, assume essentially no change
- In instances where a channel is dropped once, we just miss that measurement, basically
- All we've done is effectively missed one step
- If we calc pose w/ 8 ch., there will be a large constant offset
- John suggests using the previous 9 ch. meas. to calculate the offset, not the 8 ch.
- AKA Use last good measurement to apply correction to 8 ch. measurements
- In this case, we would not miss a step
- In the case of an 8 ch., just use the last good 9 ch. (or last two good 9 ch.'s)
- If we are off by about 1.5 deg. C, the Ciddor correction becomes very important (for 0.5 deg C, not as much of a problem)
- Have to be careful when experiencing large Jacobian changes


Yang and Andrew's Discussion:

- Evaluating image stability best done with WFS data

John's Evaluation of the Tool:
- Some non-physical telescope optical phenomenon observed on sky (2018)


Engineering Time w/ LBC on Sunday:
- Have FPIA and LBC running
- Andrew suggests we would like to test channel dropping before that
- Yang not able to make the time, lots to work on
- Andrew suggests using guide cam images, centroiding, and comparing these with what our measurements would drive mirror to
- However, not able to differentiate the actual telescope tracking, atmospheric image motion from pose changes
- We can show the non-existence of [some sort of problem]
- Can apply difference in channel number to the pose reconstruction after the fact, with a real star
- Normally, when running LBC, guide camera images go to DMS and centroiding is performed
- Sorry, the guide camera corrections are stored on DMS forever, not the guide camera images
- Just need to know the data range to look at the guide camera correction data
- So what to do with Sunday's engineering time...
- Yang suggests passively collecting TMS data for offline analysis
- There are some LBC image issues, and the engineering team is wondering if TMS team can help
- Apparently there are some ongoing LBC image problems which need solving
- Yang also mentioned FPIA problems... we want to fix this w/ our intra/extra focal method

-- TrentonBrendel - 02 Oct 2020
Topic revision: r1 - 02 Oct 2020, 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