VertexRSI Antenna 9001 Optical Pointing Acceptance Telecon Minutes


TIP Last Update: JeffMangum - 06 January 2009

Telecon 2008/08/22 19:00 UT Regarding GET_xx_ENC/xx_POSN_RSP Confusion, Metrology Latency, and GET_METR_DELTAS

  • Participants:
    • Brian Glendenning (ALMA CIPT)
    • Ralph Marson (ALMA CIPT)
    • Jorge Ibsen (ALMA CIPT)
    • Lewis Knee (AIV)
    • Art Symmes (ALMA AIPT)
    • Jeff Mangum (ALMA SIPT)
  • Differences Between GET_xx_ENC and xx_POSN_RSP:
    • LK presented results from his four test measurements done night of 2008/08/21-22 which were designed to expose differences between the use of GET_xx_ENC and xx_POSN_RSP (see COMP-2412 for full details)
      • Test 1: GET_xx_ENC and metrology OFF
        • Free fit to 14 terms gave a model similar to the default (input) model.
        • RMS = 1.16 arcsec
        • Constrained fit using default (input) pointing model terms produced RMS = 1.34 arcsec.
      • Test 2: GET_xx_ENC and metrology ON
        • Free fit to 14 terms gave a model similar to the default (input) model with (marginally) higher scatter.
        • RMS = 1.23 arcsec
        • Constrained fit using default (input) pointing model terms produced RMS = 1.63 arcsec. Clearly larger scatter due to inappropriate handling of metrology when using GET_xx_ENC.
      • Test 3: xx_POSN_RSP and metrology OFF
        • Free fit to 14 terms gave a model which looks like a differential model with respect to the default model.
        • RMS = 1.29 arcsec
      • Test 4: xx_POSN_RSP and metrology ON
        • Free fit to 14 terms gave a model in which many terms were close to zero, but some were not, and the scatter was significantly larger.
        • RMS = 2.01 arcsec
        • Three terms are discrepant (i.e. not zero) in this fit:
          • NPAE: -33.70+-4.758
          • AN: -5.47+-0.311
          • AW: +34.18+-0.327
        • The AN and AW terms are approximately equal to their input model values (-1.7 and +40.4, respectively).
        • Do not currently understand aberrant NPAE term.
      • Note that during these tests, ACU pointing model terms were all set to zero (including extra contractor terms designed to hold tiltmeter zero-point offsets).
    • Concluded that Computing should make change to use xx_POSN_RSP rather than GET_xx_ENC (internal variables which track these monitor points are actPostionxx and encPositionxx, respectively). ACTION: Computing (Marson/Ibsen) will effect changes necessary.
      • UPDATE: RM/JI installed software changes and LN tested. Results are:
        • Apparently there is a sign error in the change which included the input pointing model correction.
        • Aberrant HECE/HESE terms result. Perhaps refraction is not properly being accounted for?
        • Computing has been asked to work on modifications (with AIV oversight).
  • Proposed Changes to GET_METR_DELTAS Monitor Point:
    • Recall that at yesterday's meeting KW pointed out that GET_METR_DELTAS monitor point delivers all corrections, including SPEM and metrology, which is furthermore not synchronized to TE.
    • All agreed that contractor(s) need to change information delivered on GET_METR_DELTAS monitor point to include only metrology corrections.
    • No need to synchronize with TE.
    • ACTION: BG (Computing) will work with the antenna contractors to change the definition of this monitor point to include only metrology system corrections. Synchronization to TE is not needed.
  • Usefulness of Measurements Made To-Date With/Without Metrology Active:
    • Measurements made with GET_xx_ENC and no metrology are good non-metrology pointing measurements.
    • Measurements made with GET_xx_ENC and metrology active likely have an additional error due to improper handling of metrology.
  • Follow-up meeting will be called as needed (early next week).

-- JeffMangum - 23 Aug 2008

Telecon 2008/08/21 17:00 UT Regarding GET_xx_ENC/xx_POSN_RSP Confusion

  • Participants:
    • Brian Glendenning (ALMA CIPT)
    • Ralph Marson (ALMA CIPT)
    • Joe McMullin (AIV)
    • Lewis Knee (AIV)
    • Art Symmes (ALMA AIPT)
    • Klaus Willmeroth (VA)
    • Jeff Mangum (ALMA SIPT)
  • Differences Between GET_xx_ENC and xx_POSN_RSP:
    • JM asked KW to explain the difference between information delivered by GET_xx_ENC and xx_POSN_RSP:
      • GET_xx_ENC:
        • Raw readout of encoder (in bits).
        • No ACU-based corrections (pointing model, metrology, or encoder offset) are applied.
      • xx_POSN_RSP:
        • Actual position of the antenna.
        • Calibrated (in turns) encoder position which includes all corrections (ACU-based pointing model plus metrology) that are active (i.e. antenna is in autonomous mode).
      • If one is in encoder mode, the information delivered by GET_xx_ENC and xx_POSN_RSP are identical (after calibrating GET_xx_ENC from bits to turns and accounting for encoder offsets).
      • KW recommends that ALMA always use xx_POSN_RSP.
  • Access or Derivation of ACU-Based Pointing Model Correction:
    • RM asked KW how to get at total ACU-based pointing model (SPEM in the VA parlance) from a single or several monitor points.
    • KW pointed out that GET_METR_DELTAS monitor point delivers all corrections, including SPEM and metrology. ACTION: KW will verify this fact with VA computing.
      • UPDATE: KW provided the following information on 2008/08/22
        • GET_METR_DELTAS
          • includes corrections based on SPEM, tiltmeters and linear sensors
          • includes only those corrections that are currently switched on
          • is not synchronized to the timing pulse
    • BG and JM noted that this usage seems inconsistent with the name of the monitor point.
    • RM suggested that if GET_METR_DELTAS could be modified to deliver only metrology-based corrections (and not SPEM-based corrections) then ALMA could use this monitor point, in combination with xx_POSN_RSP and GET_xx_ENC to isolate any SPEM corrections applied in ACU.
    • BG noted that this change should be coordinated with other antenna vendors. ACTION: BG will coordinate change through ICD development involving Computing IPT.
      • UPDATE: BG (Computing) will work with the antenna contractors to change the definition of this monitor point to include only metrology system corrections which are synchronized to a TE.
  • Confusion Regarding Measurements Made To-Date With/Without Metrology Active:
    • RM confirmed state of software as of last night (2008/08/20). Current software uses GET_xx_ENC for position response from ACU.
    • LN and JMc noted that all-sky optical pointing measurements that they have made suggest that metrology is being accounted for in our derivation of the pointing results.
    • JM noted that LN and JMc results are inconsistent with our understanding, and VA's description, of the differences between GET_xx_ENC and xx_POSN_RSP.
    • JM also noted that LN and JMc results are inconsistent with test results done by Nicholas Emerson and Tzu.
    • At this point all expressed confusion regarding this inconsistency.
    • JMc proposed that LN do the following set of tests on the night of 2008/08/21 to isolate the behaviour of the GET_xx_ENC and xx_POSN_RSP monitor points:
      • Make all-sky optical pointing run...
        • with GET_xx_ENC and metrology off.
        • with GET_xx_ENC and metrology on.
        • with xx_POSN_RSP and metrology off.
        • with xx_POSN_RSP and metrology on.
    • JM later (after the telecon ended) suggested that if metrology could be made to "misbehave", thus making the difference between metrology-on and metrology-off tests more apparent, then the results of these tests could be more easily interpreted.
    • JMc noted that results and data for these tests will be posted to AIVMAN-505.
    • JM proposed that we have a follow-up meeting tomorrow (2008/08/22) after results of additional optical pointing measurements have been made. Follow-up meeting will be held 19:00 UT on 2008/08/22.

-- JeffMangum - 21 Aug 2008
Topic revision: r1 - 2009-02-06, JeffMangum
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 NRAO Public Wiki? Send feedback