Proposed S1 Calibration Schedule
                For LLO 4k and LHO 2k/4k IFO's
                    M. Landry June 20, 02
                --------------------------------
 

1) Twice-daily swept sines with the autocalibrator (one daytime (assuming
locking is possible), one night time).  Each of the two sweeps should only
cause about 5m of ifo downtime each, that is, time in which the physics mode
button will be off.  This daily downtime is of the same order as that of
relocking, so, an insignificant hit to the duty cycle.

2) Two calibration lines injected per interferometer, for the entire duration
of the run, one below the unity gain point, one above.  These lines will
allow the tracking of the stability of the calibration for the duration of
the run.  In principle, one line will do, however, a second will alert us to
online changes in the shape of the transfer function (daily swept sines will
give statistics on this too).  AWG excitations have proved unstable, hence
we have to be more careful for S1.  Options include

            a) AWG excitations would be monitored by some DMT process that
               alarms, restarts excitations when they vanish, or
            b) excitations injected via hardware (noise considerations would
               have to be addressed **NOTE Rana prefers this greatly, in
               part due to the impact on the LSC FE cpu of testpoints**), or
            c) LSC FE code change to sum excitations in digitally

3) Two dedicated calibration runs to accumulate more data such as sweeps,
loop gains and control signal calibrations.  The two runs would be just
prior and just after S1:

      Dedicated Calibration run #1: Sat. Jun 29 4am-8am PDT
      Dedicated Calibration run #2: Mon. Jul 15 8am-noon PDT

Note this latter time may have to be pushed back until Monday evening at LLO,
and quite possibly LHO too.  Or, we could grab time from the S1 run Monday
morning, e.g. 4am-8am Mon 15th.

Regarding the excitations (both discrete and swept sine): we have to agree on
a way to store the waveform.  Given our current configuration, the only way
to do this for S1 is to write a dedicated RDS to disk.  Furthermore, note
that the IFO state vector currently ignores an excitation on the ETMX, so
we can inject calibration lines.  The waveform should thus only comprise of
the two discrete calibration lines.  We should have some method of checking
no other waveform has been applied (DMT software?  shift check?).

This information may also be reached at
 
    http://blue.ligo-wa.caltech.edu/engrun/Calib_Home/
 
 

 -- 
Michael Landry
LIGO Hanford Observatory
P.O. Box 159
Richland, WA 99352

phone: 509-372-8133
fax:   509-372-8137
http://www.ligo.caltech.edu