Attached shows the comparison of ASC-POP_A_NSUM (one of the QPD on the sled in HAM3) and LSC-POP_A (LSC sensor in HAM1) before (references in DTT) and after (current) the PR3 realignment, both in full lock. To the left is the trend showing the PR_GAIN (which comes from LSC POP_A in HAM1) at the top. I chose 2023/11/30 070000 UTC as pre-adjustment and 2023/12/02 090000 UTC as post-. Note that Sheila has made yet another adjustment after this "post-adjustment" time, but that's not included in this alog.
If you compare ASC-POP before and after (green and red on the right bottom) the noise level was about the same. If you compare LSC-POP before and after (brown and red on the right bottom), clearly there was a lot of extra noise before the realignment. This suggests that there was extra clipping in the LSC POP path. After the adjustment, ASC and LSC POP agreed with each other as you can see from the coherence (red trace on the right top), so the extra clipping in the LSC POP path was gone. Note, that does not necessarily mean that there is no clipping inside PRC which will appear as common clipping on both ASC and LSC POP sensors.
For the difference of these paths in HAM3, see https://dcc.ligo.org/LIGO-D1000339. LSC beam is the PR2 transmission of the back-propagating beam (coming from PR3), and ASC beam is the transmission of the forward-propagating beam (coming from PRM). The only common component in these paths is the fixed 3" optics right behind the PR2.
In the 1st attachment left, I made a comparison between 2023/12/02 0900 UTC traces in the parent alog (i.e. after Sheila's 1st adjustment but before the 2nd adjustment) and now (after Sheila's 2nd adjustment but before Jenne/Naoki started moving PR2 further). There's a difference between now and 2023/12/02 noise level, but no indication of extra clipping in the LSC POP path.
Then I looked further back in time where it was running with 160Mpc (2023/11/28 9:30 UTC) in the attached right. Orange/Cyan in the right plot are the same as orange/cyan in the left (i.e. these traces represent now). Pink/black represent the good old time. The LSC pop (black) was much noisier than ASC POP (pink) back then but this isn't the case for ASC POP. This seems to mean that the extra clipping was present when we were running at 160Mpc. If you look at the ASC POP now (orange) and back then, it looks somewhat noisier now below 6Hz or so.
In the second attachment is the comparison between bad time (2023/11/30 0700 UTC before Sheila adjusted PR3) and the good old 160Mpc time. Bad time was worse.
To summarize,
- there was an extra clipping in the LSC POP path when things were great (160Mpc).
- extra clipping seemed to be worse on Nov/30 than in the great days.
- extra clipping seemed to be better than in the great days after Sheila made an adjustment (Dec/02) to the point where you cannot see extra clipping.
- extra clipping may be somewhat better after Sheila's 2nd adjustment (today) but hard to tell.
- hard to quantitatively assess the BNS range difference between now and the great days due to useism difference.
If you want to use the same dtt template, use
/ligo/home/keita.kawabe/20231204_POP_clipping.xml
Everything is already saved in the references, so you can just run the template.
