[evlatests] Tardy Subreflectors...

[Rick Perley]

     This is a summary of the 'tardy subreflector' problem.

     Three antennas show delayed response of the subreflector focus to a 
change in band.  This result is based on two observational runs, using 
identical SBs, other than the start time/date.  The effect appears to be 
intimately connected to observations at X-band.  See below for particulars.

     For the three antennas in question, the effect shows as follows:

     a) A change of band is requested.

     b) Most antennas' subreflectors move to the appropriate new focus 
position with a few seconds.

     c) For three antennas, the subreflector movement often starts 
late.  The delay can be anywhere from 20 seconds to (likely) infinity.  
(This means that for many of the scans, the subreflector hadn't started 
its movement by the time the scan ended).

     d) The problem is not one of slow motion, as the phase plots 
(attached) clearly show the movement starts and ends over the normal 
time period (~5 seconds).

     e) The system is aware of the problem, in that the data are flagged 
until the moment when the subreflector reached the correct position.

     f) For antennas ea13 and ea26, the effect only occurs when moving 
from Ka band to X band.  For this band, the problem is very common -- 
about 75% of the data from these antennas is flagged bad.

     g) For antenna ea05, the effect only occurs when moving from X band 
to Ku band.  At least half the scans are affected.

     h) No antennas show the effect when moving from Ku to Ka, or Ka to 
K, or K to Ka bands.  (These are the only combinations tested in the 
script).   The top panel shows an antenna whose subreflector moved 
correctly.  (Note the small scale).  The bottom two panels show ea13 and 
26.  Note that the motion began at different times, but completed within 
the usual time frame (for 7 mm motion).  Due to the small motion, no 
discernible change in amplitude is seen.  (Which raises the question of 
why are we moving the subreflector at all ...).

     i) Note that the subreflector *rotation* command was executed in 
the normal time.  This is a *focus* issue only.




-------------- next part --------------
A non-text attachment was scrubbed...
Name: LATE-SUBR
Type: application/postscript
Size: 36104 bytes
Desc: not available
URL: <http://listmgr.nrao.edu/pipermail/evlatests/attachments/20180425/8d330403/attachment-0001.ai>