Summary of TMA tests#

Note

This technote is a work-in-progress.

Abstract

This technote is linked with: https://rubinobs.atlassian.net/browse/SITCOM-1254

This technote sumamrises the tests (both past and ongoing) verifcation tests carried out with different operation al components of the TMA. With the analysis still in progress we are expected to derive conclusive behaviour of the TMA and populate this technote with summary of each tests, inferences, and proposed next steps.

This technote also serves as an index to the other tech notes produced on each test or data analysis for TMA performace.

Introduction#

Most of the performance analysis is based on the data from 1-3 nights in the second half of March 2023, except for the slew analysis which also included the data taken since November 2022.

TMA Performance Analysis#

Index of tests and verification#

Data taken on

Analysis finalised on

Results presented in
26 January 2023
03 February 2023

SITCOMTN-57
09 March 2023
14 March 2023

SITCOMTN-063

SITCOMTN-64
09, 17, 22 March 2023
23 March 2023

SITCOMTN-065
15-18 March 2023
21-23 March 2023
27 March 2023

SITCOMTN-066
26 January 2023
23 March 2023
24 March 2023

SITCOMTN-067
01 November 2022 -
30 March 2023
14 April 2023

SITCOMTN-068
21 March 2023
05 April 2023

SITCOMTN-071
21, 22 March 2023
14 April 2023

SITCOMTN-73
09, 24 March 2023
04 July 2023

SITCOMTN-77
29-30 June 2023
16 August 2023

SITCOMTN-80
20 February 2024

SITCOMTN-110
23 November 2023
21 April 2024
15 May 2024

SITCOMTN-112

Pointing verification#

  • SITCOM-1310 While the encoder tracking errors are small, we discover significant tracking drfit aling the N-S direction in both Az and El axes (*pending verification*).

  • SITCOMTN-073 Based on data limited to two nights, variation was presented in relative pointing for RA – in both forward- and backward-commanded pointing – spanning the range of -100< \({\Delta}RA\) < 100 arcsec.

  • SITCOMTN-077 The drift during tracking was analysed on data of 1 night and was observed to be of order of \(10^{-4}\) arcsec for both RA and Dec. The average drift velocities for both RA and Dec were similar in magnitude but different in directions. The average angular drift velocity (0.6 arcsec/min) is within 1sigma of the expected. Hence, from this dataset we conclude a drift behavior of the tracking within normal expectations.

  • SITCOMTN-063, SITCOMTN-065 Under a 3.5 degree random offset, it was clear that the pointing model (for a test position set 1039 to 1071) progressively degraded with time with an average offset rate of 0.55 arcsec /min. Correcting for this rate, we get an on-sky precision (1 \({\sigma}\)) of 0.3 arcsec for RA and 0.21 arcsec for Dec. This is close enough to the required 0.2 arcsec precision.

TMA jitter verification#

  • SITCOMTN-112 This analyzes the TMA jitter and slew and settle performance over a range of dates. Over 91% of the slews meet the <0.01 arcsecond jitter specification. Many of the ones that don’t meet the specification are caused by deficiencies in the way we archive the data, not by true jitter failures.

Slew and settle verification#

  • SITCOMTN-067, SITCOMTN-068 The velocity, acceleration and jerk of TMA was analysed under 3.5 degree random slews. Using a methodology that involved identifying the slew profiles, and using spline interpolation, it was determined all the slew profiles met the specifications without hitting the maximum permissible design limits. It was also noted that majority of the flagged/extreme slews could be corrected through a smoothing algorithm applied to the motion profiles to account for plausible noise introduced by the encoder (which prevented obtaining accurate slew profiles originally). Furthermore, slew and settle requirements for 3.5 degree slews with elevation less than 60 degrees are met 96% of the time. The failures tend to occur at times with larger slews in elevation relative to azimuth.

  • SITCOMTN-112 This analyzes the TMA jitter and slew and settle performance over a range of dates. 95% of the slews less than 3.5 degrees meet the 4 second specificatuon.

StarTracker#

  • SITCOMTN-071 The current pinting pointig model devloped for TMA uses the narrow camera. The offset from the narrow camera to the fast camera is calculated and verified, such that when the TMA is pointed at an object, it will be positioned at the boresight of the narrow camera, and then introduce a known offset to put the object in the field of view of the narrow camera. It was found that to center the object in the fast camera, we should slew to the object and then apply the following command: mtcs.offset_azel(az=-208, el=508). However, it was also found empirically on multiple nights that the sign of the az offset was incorrect, and the command mtcs.offset_azel(az=208, el=508) successfully put the object in the fast camera field of view, although not always centered.

Others#

  • SITCOMTN-066 We have not resolved encoder disagreement events that have been noted in data from 2023. More data and detailed analysis is required to understand patterns and potential cause of such disagreements.

  • SITCOMTN-080 Torque hysteresis was also observed during the first TMA balancing post addition of M1M3 cell - this behaviour was observed at a minimal of 1% of maximum speed and at a very specific elevation angles of < 3.85 degrees. A detailed data analysis of of historical slews identified this behaviour to be present since before the first TMA balancing. A similar elevation torque anomaly also occurs at zenith. The possible causes for such strain in required torque (or drag) are linked to missing elevation structure magnets in the axis motor, disabled elevation drives (2 at the time of observations), elevation breaks and elevation axis hard stops.

Pending verifications#

  • SITCOM-704 - First pointing model generation-Data acquisition preparation

  • SITCOM-1120- TMA Brake Analysis by measuring the distance the telescope moves after initiating e-stop

  • SITCOM-1223 - TMA Capacitor Bank discharge vs. Acceleration profiles (SITCOMTN-110; SITCOMTN-123)

  • SITCOM-1241 - Confirm Slew and Settle time using Fast Star Tracker

For all the tests, the requirements for TMA are extracted from <link>