TGBT15A_915_92 - Test Stokes U and V with new VEGAS bof file

Goals

• Test behavior of Stokes U and V data with a new BOF file that uses registers in place of delay elements.

Details

Session begins at 10:00

• Switched VEGAS to pulsar branch
• Will use 00-CoherentModeArtifactTesting in c1500x0512 mode (this is the mode the BOF file was built for). Will set-up at S-Band to try to avoid FFT overflow issues.
• Taking one scan with current BOF file (c1500x0512_x14_7_2016_Feb_24_1437.bof) for comparison
  • SB submitted at 10:06
  • Scan is #1
  • SB ended at 10:08
  • Accidentally had vegas.fold_dumptime set to 1.0 seconds, so there are lots of sub-integrations. Data should be fine but going to take another scan anyway
• Taking second scan with current BOF file, 10-s dump times.
  • SB submitted at 10:10
  • Scan is #2
  • SB ended at 10:11
  • Data don't look great. It's almost like there are drop-outs in the phase domain. Dropped packets? Something wrong in the PFB? Definitely seeing phase wraps in U/V, though.
• Switching to test BOF file (c1500x0512_x14_7_rtd_2017_Mar_21_1121.bof)
  • SB submitted at 10:22
  • Scan is #3
  • First attempt failed to start HPC subprocess and aborted. Maybe a systemstop/start is needed.
  • Resubmitted at 10:26
  • Scan is #4
  • SB ended at 10:28
  • No obvious changes in U/V data
• Going to take a scan using test BOF file but at L-Band. Will try changing fftshift to all f's and set vegas.scale to 110
  • SB submitted at 10:37
  • Scan is #5
  • SB ended at 10:39
  • We are getting dropped packets. Possible source of the features in the phase-domain?
• Taking one more scan to look at dropped packets
  • SB submitted at 10:45
  • Scan is #6
  • SB ended at
  • Seem to have consistent 30% dropped packet rate
• Switched to release mode at 10:48
• Taking verification scan
  • Using CheckMode1 scheduling block
  • SB submitted at 10:52
  • Scan is #7
  • SB finished at 10:52
  • Data look good

Session ended at 10:54

Conclusions

Phase wrap in Stokes U and V data still seems to be present with new BOF file, but dropped packets could complicate interpretation.

-- RyanLynch - 2017-03-22