Jenne, Keita, Louis, Elenna

As noted in alog 91610, we struggled to use the green baffle scripts for setting the pointing of the ITMs.  We ended up doing the process of the baffle script by hand to set both ITMX and ITMY.  We did so with the IMC unlocked. 

I think the issue (or, at least a significant piece of it) is that the script is likely looking for the _NORMALIZED_ channel to have a peak at 1, but actually it's getting up to 10 or 12.  Also, at least with our (mis)alignment yesterday, the main beam was getting into the baffle PDs enough to send their values above 1, even without any actual green beam being on the PDs.  So, the script was "optimizing" the "green pointing" by looking at IR beam that was completely independent of any ITM pointing.

As I write all of this, I realize that I did not personally run the ISC_LOCK DOWN and SDF_REVERT states, so did not triple check that all the settings (eg PD gains) were set correctly.  We should check that before making modifications to the script.

Fix #1: We set the IMC to OFFLINE, so that we wouldn't get any IMC flashes.  This removed the IR beam from the equation, and the baseline level on the ETM baffle PDs went down to (or below) 0, rather than always being above 1.

Issue #2: Even with the IMC offline, the script would think things were close enough when the baffle PDs read a value above 1, even though that's actually on the edge of the PDs (whos peak values were closer to 10).  To solve this, we just did the optimization by hand (put the sliders to roughly the right place, engage the M0_TEST offsets with the appropriate values, then adjust the sliders to find the peak of the PDs).  We then noted the slider values, then put in the opposite TEST offsets for the other PD, and then re-adjusted the sliders to find the best place on the second PD.  There are a few options for fixing this for use with the script.  We can take the time to check that the gains are set correctly, such that the normalized value really should be going to 1.  Or, we can ask the script to look at the rough peak heights, and then use that as the peak value to be looking for before considering it optimized. This second option is certainly less reliable than the first option of just fixing the normalizations. 

Issue #3: For some reason that I don't understand, the values for the M0_TEST offsets, when applied to a well-aligned ITM, don't put you onto the ETM baffle PDs.  You still have to move the sliders.  I think this is equivalent to saying that when we applied those M0_TEST offsets and then optimized for the two baffles, we expect that the resultant slider values should be similar but they were not.  We averaged our resultant slider values, and used those for the ITM positions.  Someone should work to understand why these offsets aren't the correct values - has something changed?  Or maybe there was a setting somewhere mis-set?

Issue #4: I don't think that the baffle PD scripts have a catch point for ctrl-c and a graceful exit.  While we were running the script one of the times, we had a power glitch and tripped SEI and SUS at both end stations.  I ctrl-c'd, but that did not stop the excitations. Louis "took over" the EXC slots using awggui, and that successfully stopped the excitations.  We could have also killed the excitations.  Either way, it would be nice if the script was graceful about it (turn off excitations, put in the misalign offsets that gets done at the end of the script, set sliders back to original values).

Issue #5: A few times the script noted while it was doing things that it failed to find one of the baffle PDs.  That happens, not a big deal, and great that the script says so.  But, if that happens, it would be nice if the "the script has finished" message adds information about whether or not it successfully found both baffle PDs.