Dan, Kiwamu, Sheila

Today we made several gaurdian changes:

• LSC_CONFIGS is now called ALIGN_IFO, and many states that were not necessary for inital alingment are gone.  I've moved the PRMI stuff to another file called PRMI.py, although this isn't a working guardian at the moment. The plan is to eventually merge ISC_DOF and LSC_CONFIGS
• We are now engaging all the DRMI ASC loops within a few seconds, and we are offloading it durring each lock (in the state TR_CARM).  It seems like this offloading has helped reduce our time to relock DRMI.  
• The CARM offset reduction is faster by about 40 seconds than it was. 
• We are now engaging ETM violin mode damping and ITMX bounce mode damping in addition to the ETMY bounce mode damping we had previously included for each lock.  
• THE ALS VCOs are now parked each lock, in the state BOUNCE_VIOLIN_DAMPING

I looked for a signal from the QPDs which was insensitive to TMS angle.  For the X arm pitch I used TRX A - 0.5 TRB and for Y arm pitch 1TRY_A-0.65 TRY_B seemed good.  I closed a verry low bandwidth loop to ITMX using this signal, screen shot is attached.  

We found that we could increase the recycling gain from ~27 to ~33 by moving the ITMs with ASC lops closed on IM4, PR2, BS, common ETMs, and differential ETMs.  We had to move ITMX by as much as 5 urad to do this, and we found that we would loose the lock soon after getting a good recycling gain.  Kiwamu calcualted that by moving ITMX 5 urad, we are moving the spot on the test masses by 1 cm assuming that the ETM is controlled. This may be due to a problem with the DHARD loop. 

We also spent some time looking for a Transmon QPD signal that would be insensitive to motion of the transmon.  We closed a loop to ITMX pitch using TRX_A - 0.5 TRX_B, the attached screenshot shows the settings.  This loop was verry low bandwidth.