2022 Jul 08 - TCS3 Engineering (T.Denault, G.Engh ) Goals TCS3 pointing run. (1st half of night, Fri 07/08/2020, 9:05pm-02:30am) Last one was done on 2019-04-09. 1. Encoder Measurement * DID NOT: Asked DayCrew to clean/inspect gears/encoders/tachometer. Did 2 encoder measurement recently, 6/20 (after encoder repair) and 7/7 (day before pointing run). Decide to use the 6/20 encoder ratios since there was lots of daytime observering with Texes in July. New ratios from r-220620.dat are 19.9529791343 19.9279676281 Ref: data/encoders/ 2. Setup. How is the tcs servo performance? Started early at 7:50pm (Thanks to Greg for starting his shift early) I noticed the software DEC limits are -50 60, not the -55 67 prior to Texes. Why? - Spex is in the center. - Collimate and balance are set. Hexapod is used. Tracking is OK. Following error range is HA 0.121, Dec 0.083. 220708-tcs-tracking.png 3. Benchmark current TCS Using current TCS version v220531. Pattern: benchmark13 tpoint file: b1.dat .ape file: b1.ape Takes longer due to slow slewing speeds. Done at 8:50pm 4. Pointing Run Change TCS to new version v220707 (with new encoder ratios). Starting at 9:05 - using GRID w/ 70 stars Done taking data at 10:35 .. 1hr 30min (25 mins longer that last time). tpoint file: point.dat .ape file: point.ape used all data, RMS = 92.97 Removed #1, RMS = 14.35 Removed #6, RMS = 13.65 Removed #6, RMS = 12.95 Adjusted data file is: point-adj.dat tpoint.log - log of fit using point-adj.dat point.ps point.pdf - tpoint graphs (postscript and PDF) point_coeff.txt - resulting coefficients. Updated .tcs-init with new coefficients # coefficients from 2022/07/08 pointing run: IH +70.5444 ID -268.7688 NP -8.6522 CH +14.0633 ME +12.8333 MA +22.2729 HCES +64.3070 HCEC -49.0237 DCES -412.9225 DCEC +115.9038 FO +35.4594 TF -66.5536 TX -27.3690 5. B2 benchmark Doing benchmark "b2" with new coefficients Started at 11:16... finished at 9:22 tpoint file: b2.dat .ape file: b2.ape Now, let's compare the 2 benchmarks: Here is a graph of HA error of b1.dat vs b2.dat gnuplot to make the graphs: plot 'b1.ape' using 10 with linespoints title 'b1 HA pt.madj' replot 'b2.ape' using 10 with linespoints title 'b2 HA pt.madj' Here is a graph of DEC error of b1.dat vs b2.dat gnuplot to make the graphs: plot 'b1.ape' using 11 with linespoints title 'b1 Dec pt.madj' replot 'b2.ape' using 11 with linespoints title 'b2 Dec pt.madj' Finished at 00:05. 6. Conclusion Point run when well. Most slew target fell in the field of view. New coefficients should be an improvement of the old map. HA has a big improvement in the benchmark graph. DEC shape was about the same, but improved the offset error. Will keep the new TCS v220608 with it's new encoder ratios and pointing coefficients! Some things to think about or discuss: 1. New software limit fo the tcs? Dec range when from -55/+67 to -50/+60 2. Maybe we should adjust the number of star from 70 to 80ish? Maybe adjust the ha range from -4 to +4 to +/- 3.5? dec range from +60 to -20 to +55 to -25? 3. Data taking takes longer due to slower slew rates. Took about 4.5 hours with no problem. 4. lots of MSAR warning during slews (MSAR & Tele EL don't match). Maybe disable warning during slews?