[evlatests] Tests April 20

[Barry Clark]

1.  13 Kband was working much better than I expected.  It's somewhat
weaker than VLA antennas, but then I didn't try tweaking pointing.
I changed the delay of IF C by about 5 ns.

2.  I have created scripts sysptc2h.py and sysptx2h.py in 
/home/mchost/evla/scripts/operations
These are LST based, and match the scripts SYSPTC2H and SYSPTX2H in
the Modcomps.  Like those scripts, they can be started at any time and
do the right thing, working with the Modcomps.  I think this should be
easier than using the kenbasepoint script to make short segments of
these files.  As soon as I can get around to it, I'll also make LST based
versions of the 5h pointing scripts and the baseline script.

3.  I tried a PN3DB-like script, which didn't work.  I'll diagnose at
leisure.

4.  There was work going on on antenna 18 until about three PM.  At 
that point I tried to wake it up.  When I first looked at it, the 
T5s were showing no power in IFs A and B, reasonable power and switched
power in IFs C and D (the latter being unreasonable because we have
no modules in BD).  I reset the formatter in IF A, which resulted in
reasonable power, and a negative switched power.  The latter indicates
the antenna was out of sync, which didn't occur to me until too late.
There were no visible fringes, but Vivek looked with .4 second integrations
later, and saw very weak fringes on IF A only.  They had opposite phases 
in various integrations, surely as a consequence of being out of sync.  
The fact that IFs C and D had more or less the same total powers and 
(positive!) switched powers is suggestive of grounding problems in the 
D Rack room.  But for IF A, there is probably nothing much wrong; just
resyncing the antenna, resetting the formatter, and changing the delay
constant would probably bring it back.

5.  I worked for a while driving VLA antennas.  K band now works, and
I have a nice set of symptoms at Q band that I can debug at leisure.