[evlatests] the P band polarization saga

[Paul Harden]

On 3/12/2014 3:14 PM, Ken Sowinski wrote:
> On Wed, 12 Mar 2014, Ken Sowinski wrote:
>> So, what we really have is:
>>
>>     RCP path =~ H =~ Y
>>     LCP path =~ V =~ X

This is indeed what we verified today.

RCP = IF A/B = Horizontal dipole (P-band)
LCP = IF C/D = Vertical dipole

The entire array is currently polarized the same (consistent) per the 
above.  This is the result of having an even number of cable swaps in 
some antennas.  In most cases, we have a cable swap in the vertex room 
(that discovered today), and the feed cables in the apex have been 
swapped over time to make it right -- i.e., an even number of swaps. 
Dan nor I know how many antennas have the swapped vertex room cables.  I 
am guessing quite a few, since these heliax cables were built and 
installed during the EVLA build.  It is a pair of heliax, mismarked, 
tucked up in the cable trays, you would never suspect would have one end 
marked differently than the other.  But, easily fixed, and we now know 
how to test and suspect them.

As Dan and I frequent the array, we will keep a log of those antennas 
with the incorrectly labeled cables.  Those would also be the antennas 
with swapped feeds at the apex.  If the decision is made to correct the 
cabling, we'll do it, and we should know which antennas need the work. 
A matter of swapping cables at the apex, and vertex room, at the same 
time and verifying the polarization remains correct.  If this cable 
swapping is deemed the right path, I do recommend waiting a few months 
until we are in the D-array (at least not the A).

Thanks Rick and Ken for helping us finally figure out where this many 
years-old polarization swap has been occurring.  It's been driving us 
nuts as well.

Just 8 more receivers per antenna to check!  Can't wait.

Paul (and Dan and Greg)

PS RICK: The P-band dipole on EA22 looks about 3 degrees skewed from ideal.