[evlatests] K-band results for last Friday night

[Walter Brisken]

I ran a test at 54 different frequency pair settings, K-band continuum, staring 
at 3C48 with 0.4 second integrations last Friday night, basically looking for 
trouble.

The program was as follows : tune at just over 22 GHz and observe for 10 
minutes.  Then observe at 53 different frequencies with increasing frequency 
changes for 1 minute.  The highest frequency was just under 24 GHz.  Then 
repeat these 54 scans.

All frequency settings had fringes with amplitudes that can be explained by a 
reasonable bandpass.

No phase spins were seen.

For the 10 minute scans, I saw stable phases and no reduced-strength amplitudes 
throughout.  The only strangeness on these two scans was that all 8 IFs 
produced no fringes for the last 10 seconds of the scans. VLA-VLA baselines did 
not drop out.  This "early drop-out" feature was quite common -- always at the 
last 10 second tick within the scan.  Not all early drop-outs affected every 
IF.  With the exception of a single peculiarity (mentioned later) LL and RR 
pairs behaved qualitatively the same throughout.

Other phenomena that occured:

0.75, 0.5 and 0 strength amplitudes, sometimes occuring in two back-to-back 
integrations.  In these back-to-back cases the amplitudes were always 0.5.

"drop-ins" -- meaning the scan started with no fringes and at a 10 second event 
suddenly joined with full amplitude.  This occured one IF at a time (both LL 
and RR together though).

phase jumps during a scan of ~100 or ~200 degrees.  definitely not exactly 90 
or 180 deg.  These were fairly rare and occured one IF at a time (both LL and 
RR together).

A strange "pause" -- an early drop-out occured on IF 1&2/ant 14 simumtaneously 
with the same on IF 1/ant 16.  Then three integrations later IF 2/ant 14 
returned to normal.  The three near-zero amplitudes had phases that deviated 
just a bit (10s of degrees) from the phase before the drop-out and the phase 
returned to normal after the pause.  The odd thing is that the sense of the 
phase bump was nearly equal but opposite in RR and LL.

I have plots showing these features if anyone is interested.  Help yourself if 
I'm not in the office -- plots are on top of the file cabinet near my door.  In 
all cases IF A is in red, IF B is in blue.

-Walter