An entry was made in the SMA Operations Log by rtm on 2017 May 05 21:03:25 UTC.

( read it directly at SMA log 33429 )

Number:         33429
Author:         Taco
Categories:     General
Antennas:       none

**** Correcting the cross receiver delays for full polarization observations ****

The following should be put into the full polarization wiki page, but I think that page 
needs a substantial rewrite, and until that is done I think it's best to just have a lot 
entry for it.

When we do full polarization observations, there is an extra delay term that needs to 
be corrected. Even if the delays between all RxA receivers are correctly accounted for, 
and the delays for all RxB receivers are correctly accounted for, it is likely that 
there will be a delay difference between the entire set of RxA receivers and the entire 
set of RxB receivers. It is best if that delay is corrected before full polarization 
observations are started. Here is how to do it:

Step 1: Fix all the delays between receivers of a given set, as we normally do. 
This should be done before switching the array into full polarization mode.

Step 2: Put the array into full polarization mode by executing the command

hal9000> touch /global/configFiles/fullPolarization

Step 3: insert the appropriate 1/4 wave plates, and orient them all in the L state, 
except for antenna 1 (or whichever antenna is the lowest numbered antenna in the active 
array) which should be in state R.

Step 4: Observe a strong source, which shows clear fringes. It should be strong enough 
that it could have been used to derive the normal delay corrections. You will probably 
see big delay errors on the 1-* baselines for the RxA x RxB and RxB x RxA chunks.

Step 5: Do a newFile, and take a few scans.

Step 6: Derive the cross receiver delays by issuing the command

obscon> delayFinder.py -x {data directory name}

If these steps have been successful, you should see flat phase on the RxA x RxB and 
RxB x RxA chunks.