[evlatests] 3-bit polarization/stability test

[George Moellenbrock]

Rick-

It seems to me that isolating the time-dependence of the
samplers' cross-talk is a tricky business.   Any pol-dep _gain_
variability upstream of the samplers will masquerade
as variability in their residual poln impurity by virtue of the generalized
algebra used to describe the net gain as a downstream term.
This is why the intentionally polarizing elements of the system
(where we usually worry about leakage) are built upstream of
the amplifiers, and also one reason why the ionosphere is
especially troublesome for polarization work (even before
worrying about non-isoplanaticity).   Upstream gain variability
has to be accounted for properly---and in this case, also the
upstream leakage (from 8-bit obs, presumably)---if
the samplers effects are to be reliably isolated.   Some very
careful algebraic gymnastics will be required here...

-George


On Tue, Nov 8, 2011 at 1:52 PM, Rick Perley <rperley at nrao.edu> wrote:
>    ... was going to send this to specific individuals ... but it's
> easier to do it this way ...
>
>    We talked yesterday about a longish observation of OQ208 for the
> main purpose of determining variability of the antenna/electronics
> polarization.  There are worries about temperature sensitivities, so an
> observation of a source through dawn was felt to be optimum.  The
> previous test was of 3C286, which has the distinct disadvantage of being
> very highly polarized, making it difficult to isolate small
> antenna-based effects.   Polarization tests are much better done with a
> strong unpolarized object.
>
> Unfortunately, at this time of year, the highly preferred source -- 3C84
> = 0319+4130 (30 Jy) is an evening source, so OQ208 = J1407+2827 seems
> preferred.
>
>    But I think 3C84 should still be utilized, and the test be done in
> late afternoon-evening, when temperature differentials are nearly as
> strong as they are in the morning.  3C84 rises just before sundown.  The
> maximum rate of change of temperature in the vertex room is likely to be
> during these evening hours.  I suggest we do this *tonight*, and that we
> use Ku-band.  Amongst other things, we know the polarizer purity at
> Ku-band is better than at X-band, -- making electronics polarization
> changes even easier to detect.
>    One of the distinct advantages of 3C84 is its very high flux density
> -- 15 times that of OQ208.  This is very useful in finding other issues,
> and measuring -- for example -- the bandpass stability.
>    It is now clear at the site, and the winds are lightening up.  A
> nice calm and frosty night is forecast.  It's a good time for this test.
>
>    Rick
> _______________________________________________
> evlatests mailing list
> evlatests at listmgr.cv.nrao.edu
> http://listmgr.cv.nrao.edu/mailman/listinfo/evlatests
>
>