[daip] Cross-hand amplitude problems for BR and FD at 15 GHz
Eric Greisen
egreisen at nrao.edu
Mon Dec 2 16:03:25 EST 2013
Matt Lister wrote:
> Hi, I've been reducing BL178BI, a continuum VLBA 15 GHz data set with full
> polarization from Aug 20 2013.
>
> I've noticed that the system temperatures for BR and FD are 2-3 times higher
> than the other
> antennas. The weather was fine at both of those sites. It appears however that
> these had higher Tsys as
> well in other 15 GHz runs in July and August, so maybe this is an issue you are
> already aware of.
>
> The LL and RR amplitudes for BR and FD calibrated just fine with APCAL, and are
> consistent with the other antennas.
> However, the RL and LR cross-hand amplitudes for BR and FD are a factor of 1.5
> -2 too high. This makes the
> polarization data unusable for those antennas, as I am unaware of any AIPS task
> that can adjust the cross-hand gains
> without also altering LL and RR. The cross-hand phases look fine.
>
> You can find a UV file for one of the bright sources
> at http://www.physics.purdue.edu/~mlister/2200+420.u.2013_08_20.uvf_raw_edt
>
> You can clearly see the effect in FD by plotting amplitude vs uv distance for LR
> or RL.
>
> Other possibly useful plots are
> http://www.physics.purdue.edu/~mlister/POSSM_2.PS
> http://www.physics.purdue.edu/~mlister/POSSM_CROSS.PS
> http://www.physics.purdue.edu/~mlister/APCAL.PS
> http://www.physics.purdue.edu/~mlister/BPASS.PS
>
> Any insight you might have on the cause of this problem and whether the BR and
> FD cross-hand data are recoverable would be helpful.
I have forwarded your message to Craig where issues with the instument
are more likely to get addressed.
Have you tried simply solving for the D terms on these antennas? It is
entirely possible that whatever is causing the high Tsys might also
cause high D terms. (Something deposited on the feed horn perhaps.)
There has never been a need for a separate gain for cross-hands - that
is physically unlikely. Bad polarization separation is - alas - far
more likely.
Eric Greisen
More information about the Daip
mailing list