[evlatests] Software time March 15

[Barry Clark]

Various problems with delay setting - values seemed to change.  I've
recorded some - maybe I can make sense of it later.  In any event, 
I've concluded I could better spend my time working on a delay setting 
tool instead of continuing to try to set all 26 antennas * 4 basebands *
6 bands worth of parameters with the current primative tools.

A problem I thought I had fixed, when a script did not go away cleanly
when aborted, recurred.  This requires stopping and restarting the
Executor.  I shall have to look further for why this happened.

We tried antenna 21, and got weak fringes.  Ken suspects we may be pointed
at a sidelobe.

In the course of this, we got correlator errors due to a bad multiplier
board.  This was claimed by the correlator controler to affect antenna
21 * high numbered DCS antennas.

Tried running the system pointing file through the Executor.  Antennas
moved around as but data seemed to be flagged most of the time.  It
is likely that the antennas start their pattern at scan start, instead 
of when they arrive on source.

The problem of half of the data not making it through Idcaf occurred.
We didn't get too far diagnosing it.  At least one case was apparently
due to something wrong being sent from the Executor to the correlator
controller.  (There may well be two problems here, with slightly different
symptoms.)