[daip] More RDV22 phase cal comparisons

[Leonia Kogan]

Hi David,

PCLOD is a task which simple reformats the text file to the PC table format.
PCLOD has not used more (at least for VLBA data) because the VLBA correlator
produse the PC table on line (starting APRIL 1999(?)).

PCCOR does not make any smoothing with the input pulse cal data.
The calibrator observation is used to solve ambiguaty.
The PCCOR reads line by line the input PC table and recalculate it to the 
SN table line converting the measured image and real parts to the relevant 
delays. The phase ambiguaty is solved using the calibrator observation.

I do not understand your problem completely. So my following remark may
not have a sence for you.

As I remember I put something at the MDELAY column of the SN table and
this something is not the actual MDELAY. The measured delays between IFs
is included in single band delays. The MDELAY column is not used at the
regular(astrophisical) observations data reduction.

Do you use the MDELAY column of the output SN table?

Leonia
----- Begin Included Message -----

>From dgg at leo.gsfc.nasa.gov Mon Oct 22 15:51 MDT 2001
From: David Gordon <dgg at leo.gsfc.nasa.gov>
Subject: More RDV22 phase cal comparisons
To: cwalker at leo.gsfc.nasa.gov
Date: Mon, 22 Oct 2001 17:51:19 EDT
Cc: dgg at leo.gsfc.nasa.gov, cma at leo.gsfc.nasa.gov, pet at leo.gsfc.nasa.gov,
        dsm.weh at leo.gsfc.nasa.gov, aen at leo.gsfc.nasa.gov,
        bec at leo.gsfc.nasa.gov, afey at leo.gsfc.nasa.gov, dboboltz at usno.navy.mil,
        lkogan at zia.aoc.NRAO.EDU
X-Mailer: Elm [revision: 212.4]
Content-Type: text
Content-Length: 1212
X-Lines: 25
Status: RO

Craig,

I've put 8 more phase cal comparison plots on gemini.gsfc.nasa.gov, directory
/pub/misc/dgg/rdv22pcals. The plots with names like 'kp.mk4-vlba.ps' are 
comparisons between the multi-band delays from the MK4 correlator vs. the
raw phases from the VLBA monitor logs (the lower tone). The ones like 
'kp.aips-vlba.ps' are comparisons of the phase cals used in AIPS vs. the 
raw monitor data. It looks pretty clear that the raw data agrees with the 
Mark4, but that something is happening to it when it goes into AIPS. 

My suspicion is that some smoothing is occuring somewhere in PCLOD, PCCOR,
or CLCAL. I've always used CLCAL with INTERP 'SELF', which should not allow
any smoothing. PCCOR has some comments about averaging, but I'm not sure
what its referring to - the calibration scan or the entire dataset. Could we
get some help from the AIPS group in figuring out what PCCOR is doing? If 
there's some smoothing being done, then an option to just pass the raw 
data would be appreciated.

I'm going to look at other RDV's to see if we have the same pattern between 
the raw data and AIPS. If so, then I could generate delay corrections, and
maybe we can fix the existing databases. 


Regards,
David


----- End Included Message -----