Test new 800 MHz and 1500 MHz incoherent dedispersion BOF files. Will use J1713+0747 at L-Band. All scans will be 300 seconds long and will include cal, fold, and search mode data.
Note that an fftshift value of 55555557 will be used.
Will also get some fluxcal data on B1442+101
Will also get some coherent mode timing data on a source TBD, probably B1937+21
Details
Session begins at 08:30.
Switched vegas.conf to version with new BOF file names and fftshift value at 03:18.
Starting will 60-second fluxcal scans on B1442+101
i0800x0064 scan looks OK in VEGAS. GUPPI OBSNCHAN is still stuck at 256
i0800x0128 scan looks OK in VEGAS. GUPPI OBSNCHAN is still stuck at 256
i0800x0512 also looks OK. VEGAS levels could be a bit lower but GUPPI looks good. GUPPI OBSNCHAN is OK now.
i0800x1024 looks OK. GUPPI OBSNCHAN good here, too.
i0800x2048 looks OK in both GUPPI and VEGAS
i0800x4096 looks OK.
i1500x0064 looks OK
i1500x0128, i1500x0256, i1500x0512, i1500x1024, i1500x2048, and i1500x4096 all look OK. Levels a bit high, but acceptable, in i1500x1024, i1500x2048, and i1500x4096
Switched to J1713 and taking scans starting with i0800x0512
Most modes look OK in terms of data, but i1500x2048 mode has double cal signal and incorrect dedispersion sweep
Cal scan is also too short by a factor of 2 in i1500x2048
I noticed that the AZ, RA_STR, DEC_STR, DEC, and RA keywords in shared memory were all zeros, for i1500x0512. Did not notice if this was true in other modes.
On further examination of the relevant files, it does indeed seem like the RA and DEC keywords are zero in the fits headers for some reason. Same is true of the RA_SUB and DEC_SUB arrays in the data extension.
This definitely has not happened with previous incoherent VEGAS observations. In fact, it was correct during Friday's commissioning observations.
Moving on to coherent scans on B1937+21. Will test all coherent modes in search and fold modes for timing offsets. Will use 6-minute scans.
Restarted VEGAS in TaskMaster to see if it fixed the position shared memory parameters.
That seems to have worked. Positions are correct now.
c0800x0064 scan still shows double-peaked profile. Need to get Luke to fix this.
c1500x0064 and c1500x0128 also has multiple peaks.
Other coherent modes look OK, other than the typical dropped packet issues in the high-nchan search modes.
Session ended at 15:30. Returned vegas.conf to release version.
Conclusions
GUPPI OBSNCHAN doesn't match VEGAS in i0800x0064, i0800x0128
Need to have Luke look at double profiles in some coherent modes.