[daip] FILLM gain problems

[Eric Greisen]

Ylva Pihlstrom writes:
 > Hi,
 > 
 > Would you be able to help me out with the following problem? I am  
 > trying to load VLA L- and K-band data (from 2/21). I cannot get a  
 > decent amplitude calibration via the SETJY/CALIB/GETJY path (at L- 
 > band especially, didn't try the K-band yet). The error messages while  
 > loading the data are listed below.
 > 
 > Thanks,
 > Ylva
 > 
 > 
 > FILLM1 13:36:46 Trying to access VLA antenna gains file...
 > FILLM1 13:36:46 ZTXOP2: using translated file name =
 > FILLM1 13:36:46 ZTXOP2: /Astroapplications/AIPS/TEXT/IONS/VLA.GAINS
 > FILLM1 13:36:48 *************************************************
 > FILLM1 13:36:48 CHKORR:  4 CORR OF108 DISAGREE ON SCALING OF DATA
 > FILLM1 13:36:48 *************************************************

The current FILLM assumes that all correlators are either scaled by
the nominal sensitivity or are not scaled whatsoever.  At Q band it
turned out that the EVLA antennas were "not scaled" according to the
format but actually scaled by a nominal (fixed) nominal sensitivity
and the amplitudes were perfectly reasonable.  This is disturbing and
suggests that 1 antenna is odd - so neither all EVLA or all VLA
antennas.  Nonetheless, there should not be a problem with VLA only at
L band and I and other testers will need to see what is bothering
you.

I have included Ken in this reply since the setting of this bit is
turning into a real problem and I fear a fundamental break in the
format here.

The other messages (exponent out of range) are normal enough when an
antenna goes dead but has not yet been otherwise marked as such (the
exponent messages are suppressed when the antenna would be discarded
for any other flag condition).

Eric Greisen