[evlatests] Dropouts, continued ...

[Jim Jackson]

This is starting to sound less and less like an L302 problem and more like 
a system timing problem.  If these things worked for the first seven 
minutes then they should have stayed working.  Seeing an effect after seven 
minutes on multiple L302's in multiple antennas does not point to the L302 
as being the problem and, in fact, strongly suggests otherwise.  Multiple 
L302's should not change at the same time unless they are being commanded 
to do so or their reference is being changed in some way.  There must be 
something going on in software, the central LO system, network timing or 
the VLA system that explains this.  When these sudden changes occurred what 
was the behavior observed between the EVLA antennas?

Jim



At 04:09 PM 6/27/2006, Rick Perley wrote:

>     New firmware, which it is claimed will fix the antenna 13 dropout
>problem,
>was installed on antenna 13 this morning.  Ken and I subsequently ran a fast
>continuum observation, to see what  has changed.  Three tests were run:
>
>     A) A short run on 3C84, with 1.33 seconds integration, at X-band.
>
>     B) A much longer run on 0555+398, with 0.41 sec integration, at X-band.
>
>     C) A short run on 0555+398, at X and C bands, alternating every minute,
>with 0.4 sec integration.
>
>     Short answer:  The dropouts live on -- usually.
>
>     From test A:  Duration = 2 minutes.
>
>     No dropouts werre observed on antenna 13.
>     No dropouts were observed on antenna 18, IFs A and C.
>     Dropouts were observed on antenna 18, IFs B and D.  These occurred
>are regular 10-second intervals, 'on the 2' (02, 12, 22, 32, etc.)  They
>came in
>two amplitudes -- a drop by only a few percent, and of about 25%.
>
>     From test B:  duration of 50 minutes.
>
>    Three different behaviors were seen.
>
>     For the first 7 minutes (from 19:36 to 19:42:40 IAT):
>
>           - antenna 13 showed no dropouts
>           - antenna 18 has loads of dropouts, but only on IFs B and D.
>Three different dropout depths are seen.
>              The dropouts occured 'on the 5s'.
>           - antenna 24 was not yet on source.
>
>     There was then a 20 second gap in the data, and when the data
>returned, and entirely different family of behavior was seen over the
>next 37 minutes (19:43 to 20:20 IAT).
>
>           - antenna 13 dropped regularly on all IFs.  Only a single
>'drop depth' by 24% was seen.  All drops were 'on the 1s'.
>           - antenna 18, on IFs B and D only, continued to drop, but now
>there are two depths -- of 36%, and by 75%.  All drops are now 'on the
>1s'.
>           - antenna 24 came online half way through -- and no drops were
>seen.
>
>     Then, at 20:20, another abrupt change in behavior (with no gap in
>the data), which changed the 'drop nature' on all three dropping antennas:
>
>           - antenna 13 drops on all IFs, 'on the 4s', by 50%.
>           - antennas 18 drops on IFs B and D only, with two different
>drop amplitudes, `on the 4s'.
>           - antenna 24 drops, also on the 4s, by a factor of 4 (to 25%
>of standard amplitude).
>
>     Test C:
>
>     This was to check the phase return behavior.  In terms of drops, (at
>X-band)
>
>     - antenna 13 dropped on all four IFs, by a factor of 1.3, on the 1s.
>     - antenna 18 dropped on IFs B and D only, on the 1s, with two
>different amplitudes:  a factor of 1.3, and a factor of nearly 6.
>     - antenna 24 showed no drops at all!
>
>        At C-band, only antenna 13 gave fringest, and it dropped in the
>same manner as at X-band.
>
>     This test ran for 15 minutes, at which point Ken moved on to other
>tests.
>
>
>_______________________________________________
>evlatests mailing list
>evlatests at listmgr.cv.nrao.edu
>http://listmgr.cv.nrao.edu/mailman/listinfo/evlatests