[evlatests] Minimal Closure Errors

[Rick Perley]

    Ken wondered what the closure errors are nowadays.   The effects he 
is worrying about could produce typical closure errors of a few degrees 
and a few percent. 

    To answer this question, I observed my favorite 'dot' source, 
J0217+7349, for about ten minutes and C and X bands, in basic OSRO1 mode. 

    Ken need not worry.  Closure levels, as reported by CALIB, are 
remarkably low.  After removing delay and bandpass shapes via BPASS, a 
solution using 100 MHz of bandwidth reveals a mean closure level of 
about 0.03%.  The maximum error amongst all baselines at any one time is 
typically 0.5% and 1 degree.   A visibility plot shows nothing outside 
of the noise.