[evlatests] Strange Tcals and Switched Power behavior

[Rick Perley]

    Although the actual values of the Tcals are not critical in their 
use for gain transfer (what's important is that the noise diodes be 
*constant* and the synchronous detection measurements be reliable), 
sensible values are important if we are to utilize the derived Tsys for 
the purpose of editing or system monitoring. 

    I used BDF2AIPS to transfer data from the SDM to AIPS.  Previous 
checks have shown that BDF2AIPS is faithfully reproducing the values in 
the SDM. 

    The Saturday data show that there are numerous problems with the Tcals:

    At L-band:  Most of the values are quite nonsensical.  Typical 
values are 20 to 50 K.  Some of them are over 200K.  One antenna (24, on 
all IFs) has *negative* values of Tcal. 
    A S-band,  the antennas with receivers have at least reasonable 
values in the table.  I note that antennas without receivers have 
reasonable values also -- one wonders how this can be ...
    At C-band, three antennas, which produce decent fringes, have Tcal = 
0.000  (Antenna 14 and 23 on all IFs, and antenna 16 on B and D IFs). 
    At X-band, all values look reasonable (but are they right?) 

    I also note that antennas 10 and 24 are not detecting the 
synchronous power -- either the noise diodes are not switching, or the 
synchronous detection system is broken.  The reported values are all 
uniformly distributed about zero.  This is the case on all IFs, and at 
all bands. 

    It is important that these issues be resolved before we can usefully 
employ this new capability.