[evlatests] Regarding Vivek's odd XCF

[Walter Brisken]

The oscillatory bandpass seen on antenna 16 last week for 25 MHz and
smaller bandwidths prompted us to look at the cross-correlation (lag) 
function between 16 and a VLA antenna, and 16's autocorrelation function.  
In addition to the expected large peak near 0 delay there were two 
satellite peaks at about 320 ns (16 VLA lags for 25 MHz).  The satellite 
lines are not centered on integer VLA lags.  I estimated that the actual 
lag was 15.64 VLA samples based on the relative amplitudes seen at lags 15 
and 16 which corresponds ~313 ns.  This turns out to be 64.06 EVLA 
samples, almost exactly 2 EVLA data frames.

-W