[evlatests] EVLA antenna polarization at L, S, C, and X bands

[Rick Perley]

    The data from the 'receiver rotation' experiment are now calibrated 
and reduced.   The essential results are given below.

    L-Band. 

    Five antennas are equipped with the fully EVLA compliant receivers 
and polarizers (6, 10, 12, 20 and 26).  For these, the antenna 
polarizations easily meet the EVLA goal of 5% cross-polarization (D-term 
amplitude), across the entire 1 -- 2 GHz passband. 

    S-Band.

    Four antennas are equipped with the new-design quadrature hybrids, 
the rest are outfitted with the old design.  The new design systems 
easily meet the cross-polarization goal over all frequencies.  Antennas 
with the old design hybrids exceed the EVLA goals at the low end, from 
2000 to 2500 MHz. 

    C-Band.

    The new-design hybrids are not yet ordered, so all antennas are 
equipped with the old design.  The antenna polarizations are highest at 
the low frequency end, and exceed the EVLA goal of 5% below 5 GHz.   
Most antennas meet the 5% goal between 5 and 8 GHz. 

    X-Band. 

    Five antennas are now equipped with the new wide-band X-band systems 
(6, 7, 14, 16, and 24).  For three of these antennas, the 
cross-polarization meet the 5% goal between 8.25 and 12 GHz.  The 
remaining two exceed the level at the extreme ends (below 8.75 and above 
11.5 GHz).  The origin of this is in the wideband phase shifter. 
    The overall performance is fine, but it is interesting to note that 
the old (narrowband) receivers had far better (lower) cross-polarization 
-- typically 1 to 2% between 8 and 8.5 GHz).  The higher 
cross-polarization of these wideband receivers is the price we pay for 
full-band coverage.  As always, it is more important that the 
cross-polarization be stable than it be low, since the former permits a 
post-correlation correction.

    A full report will be issued later via the Memo series -- but must 
await Eric's return from vacation in order to sort out a few problems 
with the 'TRUEP' program.