[evlatests] Setup, not a Drift

[Rick Perley]

    Upon closer inspection, it is quite clear that the strong phase 
drift seen between the VLA and EVLA antennas is not a drift at all!  
Within each scan (and this is especially clear for the first scan, which 
lasted over 5 minutes) that there is no phase drift. 
    Somehow, upon a return to a previously-observed frequency, the EVLA 
(or VLA) phase is offset by a set amount from its last setting -- 
presumably the 180 degree phase jump is another manifestation of this.