[Evla-eng] Re: [evlatests] EVLA Cable Wraps

[Bob Broilo]

All,

> OPS that they were not safe to use until we had a chance to inspect
> them...we will use his procedure until the new proximity-sensor
> based Lead/Lag switches have been tested & installed...

This brings up an important point.

When changing arms, the Servo techs have to tell the Data
Receiver/Buffer (DRB) what arm the antenna is in.  Since we do not
move the "flipper" any more (the limits are based on fiber length and
do not change) we will need to move the Lead/Lag (flipper) switches.
These switches are neccessary because the VLA encoder is a single-turn
device, it can only determine 0-360 degrees but the VLA antenna needs
to be able to move at least 540 degrees.  The new flipper switches
won't change this, they are just more reliable and easier to set than
the old ones.

When the servo techs set the arm in the DRB it just add/subtracts a
constant degree value to the encoder position.  This is to set all the
positions in the array to a similar heading.  When this value is
applied to the enocoder, it changes the physical place in rotation
where the enocoder crosses the 0/360 degree point, which requires that
the Lead/Lag switches need to be repositioned.

This also means that we cannot always use the full potential rotation*
of the antenna.  When certain arm offsets are applied, we cross the
0/360 point again before hitting the limit.  When the antenna crosses
this point it will get "lost" until someone can go help it find itself
again.

If these arm offsets were applied at a higher level (software), then
we could set the 0/360 point to the exact center (midpoint between the
limits) of antenna rotation.  Then the Lead/Lag switches could be set
once, permanently, eliminating a potential source of error.  Most
importantly to science, it would allow us to use the full potential
antenna rotation which would minimize antennas having to turn all the
way back around when tracking an object in the sky, increasing
time-on-source.  As a bonus, it would eliminate the possibility of
driving an antenna to a "lost" state.

Since the EVLA software knows which arm the antenna is on, it could
apply this simple offset and improve the situation for servo,
astronomers, and operations.

Please consider this simple software change,
-Bob.


*(I don't have the real numbers in front of me, so this is from
memory, so apply usual NaCl).  The full potential rotation of a VLA
antenna with the fiber is (I think, Jim Ruff knows better what the
fiber is designed for) 30 to 660 degrees, or 630 degrees.  The current
rotation is limited to 88 to 632, or 544 degrees.