The 4band situation as of today:
1) Ken seems to have been wrong about what is happening to the data going through WIDAR. It seems that for each antenna there is an effective electronics gain which remains constant throughout the run. This gain seems to be due to the range of power levels being sampled by WIDAR. Thus when Ken changed the levels by hand for each antenna, the raw amplitudes changed accordingly. This must mean that any amplitude calibration differences must be due to interference which correlates from antenna to antenna or to the ionosphere.
2) The test Tracy set up could not be gotten on the telescope before Cygnus A set. Ken promises me he will get it done tomorrow.
3) The birdies seem to be coming from at least two sources. The 60, 70 and 80 MHz birdies (and any at 5 MHz intervals) seem to be coming from EVLA electronics in the antenna. The 77.25 MHz birdy is from an unknown origin. It comes and goes. However LWA does not see it. Thus it seems likely to be generated by EVLA electronics or staff somehow.
4) There are two potential improvements in observing which could be made but both require some non-standard setups.
i) Ken suggests moving the sub-band which is being used for 74 MHz away from subband 0 which is used now. This has the effect of increasing the fringe rate and thus suppressing the interference somewhat. It also gets away from from a roll-off in the bandpass amplitude at the lower end of subband 0. However, this requires had corrections to the scripts or a change to the OPT.
ii) The power range sampled by WIDAR is not ideal for 4band. It produces negative autocorrelations which apparently is due to saturation of the total power being sampled. For all programs to get the better power levels Ken would need to adjust the power levels being sampled by hand for each program in real time or he may be able to make some sort of change in the input power levels somehow which I do not understand.
Both of these changes should improve the performance but are difficult to implement with such short notice.
5) I am leaving on vacation on Thursday, so I have limited time to do anything about these problems. I will try to coordinate with Joe McMullin
about what is to be done.
6) Assuming the Tracy's test occurs it would be useful to look at it Tuesday morning if possible and give some feedback if it looks OK. As I understand it, both improvements i) and ii) will be made for that test.