VEGAS Team Meeting
March 22, 2012
- Dan Wertheimer
- Mark Wagner
- Glenn Jones
- Andrew Siemion
- Jason Ray
- Patrick Brandt
- Galen Watts
- Randy McCullough
- Marty Bloss
- Anish Roshi
No VEGAS teleconference on Monday, March 26
First on-sky observations with the full VEGAS mode-1 were conducted on the night of March 12. The first detection of PSR_B0355+54 was taken at 22 GHz, with 7.79 GHz bandwidth and 5538 channels.
A second detection, this one of PSR_B0329+54 is shown below.
Observations of a flux calibrator were executed, but it did not show up in data - plan to re-execute after verifying on/off source in Astrid. Having the calibrator as part of the observation will be useful in setting up for pulsar observation.
The current VEGAS setup is stable but running GUPPI in parallel via Astrid commands had some problems. It did record baseband data but had 'glitches that affected a small amount of data'. Fpr their observations the Berkeley team would like to change GUPPI's frequency during the VEGAS observations by are using other LOs. This worked, but it is unclear if multiple frequency changes and starting and stopping GUPPI will be interfere with VEGAS ops. It could be an artifact of changing one LO on a buss that has other LOs even though they are not changing frequency. The resolution is for now to only switch GUPPI when VEGAS is between scans.
The plan for the final GBT observation for this visit (March 22) is to rerun the flux calibration and attempt to observe a very high dispersion pulsar.
The Berkeley team plans to do remote observing when scheduled using VEGAS in its current form and switching GUPPI as noted above.
The VEGAS switching was not controlling the cal diode for the KFPA, but that has subsequently been resolved. Full end-to-end switching tests should be completed tomorrow.
The Tsys plot below shows high frequency 'noise' on two channels. These are thought to be on a second fiber IF chain and will be investigated by NRAO engineers. The plots also show a gain slope on two different channels. NRAO will investigate and see how much is being contributed by the IF modules. The IF module equalizers will possibly have to do slope correction for the system instead of just flattening the IF module response.
Berkeley ran a quick spectroscopic Allan Variance measurement of one module and it looked pretty good. Discussions around whether these should be rerun on all modules now or after more complete delivery were not conclusive. Tests require termination of the converter inputs or the results are skewed by cable reflections. There tests may be run remotely after additional thought about the procedure.
The AsiaA ADC data rate is not compatible with Mode 1/Roach 1's (16 sample/clock). This indicates the digital compatibility of these ADCs must be tested along with analog performance. The AsiaAs will be returned to Berkeley and they will start a 16->8 sample/clock yellow block converter for ongoing VEGAS Mode 1 compatibility and subsequently compile Roach 2 mode1 with that. They could re-clock the FPGA but that is not a preferred option