TGBT18B_502_14 - Timing Offset Tests with AP

Goals

• Take coherent mode data with the artificial pulsar hooked directly into the GUPPI and VEGAS ADCs to better characterize timing offsets
• AP will be set to 714 Hz. All scans will be in coherent search mode and will be 150 seconds long. We will take 4 consecutive scans in each mode.

Details

• Jason hooked things up starting at around 12:30. Had to adjust AP power to hit good input levels for VEGAS and GUPPI.
  • GUPPI is about 4 dB high in one channel, 1 dB high in another. VEGAS inputs are at -23 and -21 dB. This should be OK.
• First short scan was to check bandpass levels in incoherent mode, which look good in both GUPPI and VEGAS
• Started timing scan sequence. First scan is #2
  • Spot check of data makes it look like VEGAS scales might be low, though the AP signal is visible.
  • For some reason GUPPI is dropping a bunch of packets on GPUs 5, 6, 8, and 9 in c0800x0064 mode. There isn't any useful signal on those banks so I don't think it's an issue, but it is odd.
  • Levels definitely seem low on VEGAS. Will try again with vegas.scale increased.
• Had to play around a bunch with VEGAS levels and fftshift. Still not sure it's optimal but it should be OK for our purposes.
• Decreasing scan length to 120 seconds. First good scan is going to be #12.
  • Now we aren't getting any signal in c0800x0064 on VEGAS.
  • Same is true for c0800x0128
• Going back to original settings and just going to use what we can get. First good scan will be #17.
  • Scan #29 tried to abort but I continued on.
  • c0800x4096 was failing for some reason. CLEO message reports that VEGAS has an invalid start time? Tried doing a full system stop/start. Moving on to c1500x0064.
  • c1500x0064 also aborted.
  • Cycling VEGAS off/on. First good scan is #47.
  • Last scan will be c1500x0064, scan #61.
• Jason began connecting VEGAS back up at 15:43
• Did a quick checkout of GUPPI and VEGAS, and balancing and bandpasses looked normal.
• Session ended at 15:49

Conclusions

• Will analyze data.