[evlatests] Small frequency changes

[Ken Sowinski]

On Wed, 4 Nov 2009, Barry Clark wrote:

> With longer scans it is clear what is happening.  There is a large
> phase jump 10 seconds into the scan.  I now think that the earlier
> indication I had of a 20 second offset was being confused by the
> known bug that the first 10 or 20 seconds of a new file has wonky
> phase.

I think that the station board is making the 10 second phase jump.
It appears that at a source change, the new delays are used at the
right time, but the new frequencies needed to calculate phase begin
to be used only at the next 10 seocnd tick.  A conversation with
Bruce and Hichem is warranted.