[evlatests] P-band receiver stability

[Huib Intema (NRAO)]

Hi Rick,

Thanks for this report. It is useful to have such a long observing run,
which allows for uncovering some of the issues that are more difficult
to spot in the few-minute test observations I usually analyze.

Regarding the "two-gain-state" problem, I have seen antennas that
misbehave in one test, but not in another. We have had quite a few
issues with cables (and many thanks to the engineers that have fixed
most of them), so maybe some of those have half-broken cables. At least
both EA04 and EA16 have previously been reported as "likely
intermittent" (e.g., see WIDAR P-band status update: 01OCT2014), as well
as EA19 and EA28.

I have also noted some time-variability in the gains of some antennas. A
few also have strong bandpass ripples (like EA13-L/Y), also mostly cable
issues, which may be what is causing these gain variations. If that is
true it would be really bad, as it would NOT calibrate out (unless one
resorts to time-variable bandpass calibration, which is highly
undesirable).

Note that your antenna list for Y-problems ends with ea2, which suggests
that a number was dropped. Please confirm.

-- Huib





On 17-Oct-14 15:27, Rick Perley wrote:
>      The 30-hour 'flux models/density' run, taken last weekend, gives 
> much information on receiver stability.  Here is a short report for P-band.
>
>      In general, receiver stability is extremely good.  Most 
> antennas/IFs have constant gain to within 5% over the 30 hour period.  
> Most are significant better than this.  (And this without applying the 
> switched power).
>
>      The exceptions to this (and some are quite extreme!) are given below:
>
>      For the 'X' polarization (to be more precise, that which is 
> labelled as 'X' in the data):
>
>      ea04, ea14 and ea21.  The first is especially bad, with two gain 
> states -- one good, the other very bad.  It was in the 'good' state for 
> about 6 continuous hours.
>
>      For the 'Y' polarization, the list of misbehaving antennas is longer:
>
>      ea03, ea04, ea12, ea13, ea16, ea19, and ea2
>      ea04 is very very weak most of the time.  ea16 also has two gain 
> states, one normal, the other very not.  The rest show significant time 
> variable performance -- but sufficiently slow that it can be calibrated.
>
>      I can provide (next week) plots if these are useful.
>
>
> _______________________________________________
> evlatests mailing list
> evlatests at listmgr.nrao.edu
> https://listmgr.nrao.edu/mailman/listinfo/evlatests