All plots are at nominal. Work done on the 21st is evident in some.

Laser Status:
SysStat is good
Front End Power is -0.004753W (should be around 30 W)
HPO Output Power is 73.44W
Front End Watch is GREEN
HPO Watch is RED

PMC:
It has been locked 0 days, 0 hr 1 minutes (should be days/weeks)
Reflected power = 15.37Watts
Transmitted power = 45.19Watts
PowerSum = 60.56Watts.

FSS:
It has been locked for 0 days 0 hr and 0 min (should be days/weeks)
TPD[V] = 2.893V (min 0.9V)

ISS:
The diffracted power is around 1.4% (should be 3-5%)
Last saturation event was 0 days 0 hours and 50 minutes ago (should be days/weeks)

Possible Issues:
Front End Power is Low
ISS diffracted power is Low
LRA out of range, see SYSSTAT.adl

We were locked in DC_READOUT damping violin modes for about 4 hours today.  We lost lock due to some alignment fluctuation, and were not able to reacquire the arms via ALS for some reason.

We were having some odd behavior where the arms would acquire the correct mode, kick on the WFS 10 seconds later, then fly through resonance and drop lock, getting stuck on some higher-order mode.  (Pic 1)

Turns out that FM9 in the green WFS DOFs is a -14 dB which is supposed to be on when we first engage the WFS.  FM9 was getting turned off directly after the 10 second settling time, not after the WFS had converged.  I added a wait to prevent this from happening, ALS reacquired twice without aid.

TVo, Danny, Craig, Stefan

We made some progress damping the most offending violin modes - almost all of them on ETMX (presumably since ETMX is driven).
We kept updating the Violin wiki page: https://cdswiki.ligo-wa.caltech.edu/wiki/Violin_Mode_Table_v2

The well damping modes are: Mode 2 (507.493Hz) , Mode 6 (511.180Hz), Mode 7 (516.678Hz),  Mode 8 (507.36Hz), Mode 9 (516.781Hz), Mode 11 (508.844Hz, currently its monitor is on Mode 1)
They all have monitor channels with 20mHz BW.

We had problems with Mode 10 (510.714Hz) not convinced he is on the ETM yet...

Some additional remarks:
 - Some guardian currently keeps the ramp time to 0.2 sec. This is too short - it creates a 5Hz-wide buzz, and we see all monitor filters being kicked by it. We should go to 5sec ramp time at least.


We left the interferometer running for the night, with all violin damping settings on. Reference time for all the settings:  1219294728 (2018/08/26 04:58:30UTC). 
Attached are snap shots for the ETMX violin screen.

TJ, Aidan, TVo, Georgia, Danny 

After the ETMX ring heater test in alog 43539 we restarted the Hartmann code with the green beam shuttered. It looks like the spherical power / prism trend data and contour plotting has improved (see attached figures). The Hartmann centering on the ETM looks good in yaw but could be slightly improved in pitch. 

The guardian had trouble locking the OMC - it found the resonance, but then jumped away when trying to engage the length loop. We tracked it down to an "always on" setting in the FM2 integrator. Set it to "zero history", which fixed the problem.

ISS 2nd loop reference servo was saturating, causing the ISS 1st loop to loose lock.
We put a clear into the IMC DWON state:

  ezca['H1:PSL-ISS_SECONDLOOP_REFERENCE_SERVO_RSET'] = 2

That should fix it.

TVo, Danny, Stefan

The SRC initial alignment never worked for us, so we made the following changes:
- SRC gain up to 10000 again,
- Also engage AS_WFS centering (since we now use the AS_A WFS sensor for length locking).
- Engage both SRC1 and SRC2 at low gain, then boost the SRC1 gain by x10, so we can handle large slews across ASC-AS_C.
- Finally use convergence checkers alog the way and at the end to make sure everything worked.

We ran it three times in a row with a relatively large AS_C offset and it worked every time. Of course, the SRC still has to lock in the first place.

Brought the IFO to down, shuttered ALS X, and began an ETMX ring heater test at 1219214950 from opsws1. It will end at 1219218552.

We have a request for some DetChar help.

We would like a high resolution (better than 1mHz) resolution measurement of the frequencies of a pair of our violin modes.  They are at about 501.624-501.625 Hz and 501.628-501.629 Hz.  This will help us design damping filters for them that can be well centered on the mode we are trying to damp without ringing up the nearby neighbor mode.  It might be easier to resolve the two frequencies by finding time stretches where one mode is rung up more than the other and vice versa (like they are here)

Some times that might be useful are:

• 8-24-2018 03:09- 6UTC (DC readout)
• 8-23-2018 23:42 - 1:58 UTC 8-24-2018 UTC (DC readout)
• RF lock today in which I rang up the 501.628 Hz mode starting at 1:45 UTC on August 25th and is still locked now at almost 6 UTC. 
• Times here

Georgia and I created a violin damping guardian.  The purpose of this guardian is to remove the parts of ISC_LOCK which used to set the violin damping gains and move it into a new guardian. 

Right now the guardian is not managed by ISC_LOCK, although it should be once it has been tested more.  Things should still be carefully monitored because we are still finding settings for the violin modes. 

It has an idle state, and it has ENGAGE_DAMPING states for each test mass.  This will only turn on the settings that are currently written in lscparams for each mode, regardless of how rung up or not the mode is.  For now we have separate single optic states, because we don't yet trust all of our settings and want to watch what happens one optic at a time. 

We also have states called ENGAGE_DAMPING_RF which check the level of the monitor filter and will only turn it on if the monitor filter is above the noise level of RF DARM.  During O2 we used to always engage all the violin damping when DARM was locked on RF, but if the modes have been well damped in DC readout and not run up, then engaging the damping in this noisy state would ring them up.  We started to populate the monitor filters with 8th order butterworth bandpasses that are 20 mHz wide, but this is not completed so we can't rely on these monitors yet. 

Eventually we would also like this guardian to adjust the gains of individual modes based on how rung up they are. 

Note that this is a different guardian from the VIOLIN_DAMPER guardian that Stefan wrote to find phase settings.  

We have been steadily working through identifying and making bandpass filters for the new violin mode forest, following on from log post 43331.

I have updated the wiki with 4 new tables at the top of the page, showing currently populated filters and frequencies for each test mass. Blank filters are, as far as I know, up for grabs for new filters. I have also been putting the gains and filter values in LSCParams (as described by the "In Guardian?" column of the wiki tables), for future use with the violin mode guardian Sheila has been working on. If you populate a filter and identify a mode please update the table and LSCParams.

Modes that were identified by Jeff but still need to have their filters made are:

501.755 (IY?)

503.676 (EY?)

503.733 (EY?)

504.143 (EY?)

504.176 (EY?)

505.077 (IY?)

505.186 (IY?)

507.493 (EX?)

508.938 (EX?)

510.714 (EX?)

510.723 (EX?)

511.18 (?)

516.778 (EX?)

The first priority is identifying which mode belongs to which test mass, and setting up the bandpass and +/- 60 degree filters. Fine tuning of the gains and phases can be done using Stefan's violin mode damper guardian once the filters are in place.

The high violin modes were saturating the ASC-AS_C diode, breaking the SRC2 alignment loop.

We lowered its whitening gain from 36dB to 18dB (H1:ASC-AS_C_WHITEN_GAINSTEP from 12 to 6). Note that will mess with the ASC-AS_C calibration, so once the violins are down, we should bring it up again.

Sheila, Nutsinee Terry & Haocun

We revisited the crystal position (alog43292) as our ~22mW threshold is inconsistent with the 8mW indicated in alog41150.

We followed the procedure in alog43292: We measure nonlinear gain by injecting a seed beam, scanning the seed phase, and measuring the de-amplification and amplification of the seed.  (x=(sqrt(max./min)-1)./(1+sqrt(max./min)); gain=1./(1-x).^2;)

The plot with a sinc² fitting is attached and with a SHG fiber launch power of 18.3 +/-0.1 mW and a seed launch power of 2.7 mW the phase matching temperature seems to be ~33.8 C. The pump power is limited by the fibre power limit.

The dual resonance position was set to within about +/- 10 clicks on the translation stage driver so the temperature accuracy is still not much better than 0.1C as in alog41150.

We left the crystal at the position where we see co-resonance for 33.8 C, because we think this is about at the peak of the phase matching curve.