[evlatests] Results from the Calibrator Run 31Jan/01Feb -- P-band

[Rick Perley]

We obtained a 24-hour observation of about 20 calibrators, starting Friday morning, 31 january, and continuing through to the following day.  These sources were observed approximately nine times each in all nine frequency bands.  The total number of separate observations (source/band) exceeds 1200.  All observations are brief —less than 30 seconds.  (The target sources are strong, so SNR is not an issue).

The weather conditions were ideal — clear skies, light winds throughout.  There was a very large temperature change from night to day — -15C to +15C.

There are many goals for this long observation, one being a careful look at the performance of each band, and stressing the system to judge the reliability of our hardware and software.

------------------------------------------------------------------------------------------
This report is the first of many and focusses on the P-band results.

There were 134 observations taken of 16 sources.  Each on-source observation was 25 seconds in length.  The OPT did an excellent job in estimating the slew and setup times.

No judgement on system phase stability can be made — the phases were completely dominated by ionospheric irregularities.  For the shorter spacings (~1 Km), stability was very good — a few degrees from observation to observation.  For the longest spacings, there is no connectivity in phase between adjacent scans.  (This is not a problem for this observation ,as the target sources are all very strong, so self-calibration removes the ionospheric and instrumental changes).   There is an easily discernable difference in ionospheric phase stability bet



For 25 of the 27 working antennas, the amplitude stability was excellent — gain changes over the day are less than typically 5%, with virtually no temporal trend.  This is very impressive, given the very large ambient temperature change, and that the receivers are up in the focus box.
Two antennas 'behaved badly':
      ea02 showed gain changes of about an order of magnitude.  Both polarizations show the same gain change from scan to scan.  The amplitude data within each short scan are stable.  Phases look normal.  It appears that each observation has a completely different gain, with no relation to the previous P-band observation.  (In between two P-band observations are the other eight bands for a given source).
      ea21 in the 'Vertical' polarization showed even larger gain changes.  But the character of these are quite different.  The 'H' channel is stable.  The bad 'V' channel has the appearance of a bad connection.

The switched power ('PDif') appears to track these gain changes quite well.  However, I chose not to apply this, as the RFI environment makes nearly all the PDif values very difficult to edit and utilize.  (Although, SPW01 is nearly free of RFI, so might be usable as a proxy for all the others).  Three antennas  — ea01, ea08, and ea14 have 'zero PDif' — the switched power signal is zero-average noise.  Of these, it seems ea01 is 'zero PDif' at all bands.  The other two give good PDif values at the other observing bands, so the issue with them is unique to P-band.
The other reason for not utilizing the PDif values is the remarkably good stability in the system gains, as noted above.

RFI issues are as expected.  As the array was in A configuration, fringe winding provides fairly good defense against much RFI.  Nevertheless, the upper half of SPW02 and nearly all of SPW03 had to be flagged (250 — 270 MHz) due to some military satellites.  These are narrow-band signals (a lot of them), and very strong.  The four MUOS bands (another military satellite system), spanning the upper half of SPW09 and 2/3 of the lower part of SPW10 had to be flagged (roughly 360 to 380 MHz).  These signals are less bright, but much wider bandwidth.  Finally,  the uppermost 6 MHz of SPW 16 (474 — 480 MHz) had to be flagged, due to strong wideband signals of unknown origin.  Numerous signals of short duration, narrow bandwidth, at various frequencies in SPW12 and SPW15 had to be flagged.

With a single exception, system delays were very small (a few nsec) and stable.  The interesting exception was on ea01:  The antenna was taken out of service for a bit under two hours (10 AM through 11:54).  Upon return to service,  the delays on the A and C channels (lower half of the band) were unchanged.  But the B-channel (V polarization channel in the upper half) changed to a remarkable value of -67 nsec (!).

Bandpasses were good for all antennas.  Cross-polarization leakage was at the normal levels, corresponding to dipoles misoriented by less than 5 degrees.  I used ea23 as the reference (hoping that it is properly oriented w.r.t. the antenna frame).

Polarization calibration (using the methods laid out in one my EVLA Memos) worked very well, using 3C345 and 3C303 to determine the cross-hand phase offsets.  (DA240 cannot be used as it is heavily resolved).

One goal of this run was to judge how well our ionospheric Faraday rotation corrections are working.  For this, we needed a big change in ionospheric TEC (total electron content) between night and day.  We got that, as the active sun is doing a good job!

Overall —excellent data, which will lead to very nice results.

Rick









-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listmgr.nrao.edu/pipermail/evlatests/attachments/20250219/9278726b/attachment.html>