Laser Rangefinder Update 9/5/03 Experiment

1
Laser Rangefinder Update9/5/03 Experiment

• K. Constantikes

2
Experiment

• Designed to correlate thermal, astrometric data
  with LRF data
• 24 hours of
• Point/focus at source near NCP
• Track at 360, 38 for 2 minutes
• LRFs running continously, 15 sec sample interval
• 11 targets, some with 4 rangefinders designated
  (FA tip, el bearing)
• Zero points for each instrument
• Leakages for each instrument
• Weather data , 10 sec sample interval
• LRFs had pointing optimized for all target at
  this GBT pose

3
Range data processing

• Zero points and leakages smoothed 75 sec median
• Group refractive index smoothed 120 sec median
• LRF integration time 1 sec
• SNR gt 10 Phase estimate noise lt200 ?m
• Selected low SNR to get more range data
• RMS range residual in trilateration will ID noisy
  range data
• 97F Fem model used for range prediction and
  resolving range uncertainty
• Ranges corrected for glass offset, GRI, monument
  and rangefinder geometry
• For 4 el bearing targets (2 each side) and 2 FA
  tip targets
• Select last 60 secs of data (up to 4 ranges, 4
  rangefinders), GBT at 360, 38 ?1
• lt 650 ?m FA y, lt 250 ?m FA x,z
• lt 125 ?m el bearing x,y

4
EMS Flowgraph
5
Bounding actual FA position errors
2.4 0.8mm at FA tip
6
Bounding actual el bearing position errors
3 at el bearing 0.4mm
7
ZAG736D The best

• 1900 outage due to SW failure
• Some trend visible in X
• No data after 2130
• LRF pointing
• Dew, fog
• Expectation of 1s range error 40mm
• 3 rangefinders
• Consistent PTCS/PN/8

8
ZAG736D The best
25 minutes

• - Dispersion of range error broader due to low
  SNR threshold.
• 40mm PDOP 300mm total position error SHORT
  TERM (60 secs) ?
• Actual 1s total position error 2mm, 1s y error
  0.7mm, or 6 azimuth pointing error equiv.

9
ZEG41040L Monument errors

• No data after 2100
• Position jumps due to redundant ranging (4 LRFs)
  and monument posn errors
• Consistent with 1 mm monument posn error.

X (mm)
Number of LRFS
10
ZY107 Erratic pointing

• Very high SNR occasionally, but long outages

11
ZY102 Low laser power?

• Consistent SNR
• Degrade after sundown-
• Dew?
• Thermal pointing shift?
• SNR would be still be adequate _at_ 0.1sec
  integration
• 1900 outage was software

12
ZY103 Inadequate laser power, poor pointing

• Laser power too low for 0.1 sec integration time
• Pointing erratic too

13
Data quality statistics

• 6 Targets, 20 hours, 4 ranges per minute
• 2 retros per el bearing, 2 FA tip retros
• Possible 20x60x4 4800 trilaterations
• About half lost due to fog, dew
• Others lost to pointing/SNR issues (even though
  SNR10 threshold used)
• Total good trilaterations, percent excluding
  fog/dew
• ZAG731D .210 or 10 ZAG731P .142 or
  5 ZAG736D .611 or 25 ZAG736P .44
  or 0 ZEG41040L .241 or 10 ZEG41040R
  337 or 15

14
Reliability

• Out of six experiments, not one case where all 12
  ground LRFs were operational
• one week preparation time each case.
• Pointing problems most common, but also
• Oscillator failures
• Servo failures
• Computer (ZY and ZIY) failures

15
Fixes

• LRF outgoing and incoming beams need to be
  broadened, pointing errors identified
• Some laser diodes replaced (obsolete, will
  require mechanical retrofits
• Other parts nearing obsolesence (IRIG,
  Oscillators)
• 360 Degree retros to improve ground geometry
• Beam split and track to ameliorate detector group
  delay issues (centriod and carrier density)?
• How to ameliorate dew problem?
• More lab characterization to ensure were ready
• Major, time consuming, expensive undertaking

16
Summary

• Not only are there fundamental issues-
• Geometry and group refractive index
• And major system design issues-
• E.g., LRFs on feed arm
• But also suitability, usability, cost, and
  schedule issues.
• And we know how to fix them