I've been working on developing a method to perform the calibration swept sine measurement for the Actuators, Pcal and the overall DARM loop all at the same time. The idea is that all of the swept sines have a different start point in a united vector of frequencies, and loop to the front when necessary.

This kind of injection saves a lot of time. It cuts sweeping time down from 1.5 hours to 20 minutes for LHO; and does similarly for LLO.

This involved generating 4 pieces of code:

1. Scripts to generate well distributed frequencies, which dodge resonances that you do not want to excite; include the calibration line frequencies for either LLO or LHO; otherwise common points for both LLO and LHO; and none of the frequencies are harmonics of each other. ✓
   • Interpolate/Extrapolate amplitudes based on existing measurement parameters; and reduce UIM injection amplitudes as per my advisement [LHO aLog 65537] ✓
2. Scripts to take these frequency/amplitude vectors and the existing swept sine parameters (drive cylces, ramp times, averages, settle time, etc.) to generate settings.ini files that distribute the start points for all measurements to be run. ✓
3. Script to take in a settings.ini file through command line; run the injections; optionally record the data; optionally process injections that you have performed in the passed (currently WIP, but out of my hands); and save the data where you want to save it. ✓
4. Amendments to pyDARM to be able to digest my output file format, to produce plots/reports down the line. (about to be merged for next version). ✓

I have performed runs on test channels (CAL-CS); and injectionless recordings on the real channels (to make sure there's no memory issues). But it is important to know whether the new injections hold up against contemporary measurements in the wild.

In this aLog, I am providing a set of 5 diaggui templates and instructions on how to run the simultaneous swept sines - should Jeff Kissel find the opportunity to run them. Remember to update from the svn server to find the templates I put up. I would run the simultaneous swept sines first, as if something is to go wrong, it should go wrong here. I recommend running on a PC with at least 8 GB of RAM to be totally safe. It is supposed to take 11 seconds before it actually starts running.

Simultaneous Lines script:

/opt/rtcds/userapps/release/cal/common/guardian/
    simulLines.py

I've done the leg work in the presently undocumented series of steps it takes to prepare to run this (I guess this is my next job assuming everything is good), and the command to run it is:

    python3 simulLines.py -i settings_h1.ini -o /ligo/svncommon/CalSVN/aligocalibration/trunk/Runs/O3/H1/Measurements/FullIFOActuationTFs/

    where '/ligo/svncommon/CalSVN/aligocalibration/trunk/Runs/O3/H1/Measurements/FullIFOActuationTFs/' is the directory where the files will go - this can be altered if there is a more convenient location - just make a note so that I can find it.

Diaggui Templates (these have parameters identical to what will be run in the python script, to make proper apples-to-apples comparison):

/ligo/svncommon/CalSVN/aligocalibration/trunk/Runs/O3/H1/Measurements/FullIFOSensingTFs/
    2022-11-29_H1_DARM_OLGTF_LF_SS_5to1100Hz_20min.xml
    2022-11-29_H1_PCAL2DARMTF_LF_SS_5to1100Hz_20min.xml

/ligo/svncommon/CalSVN/aligocalibration/trunk/Runs/O3/H1/Measurements/FullIFOActuationTFs/
    2022-11-29_H1SUSETMX_L1_iEXC2DARM_20min.xml
    2022-11-29_H1SUSETMX_L2_iEXC2DARM_20min.xml
    2022-11-29_H1SUSETMX_L3_iEXC2DARM_20min.xml

I'm eager to make sure that transfer function measurements do not differ by a significant margin to those from the diaggui templates. Jeff can run this whenever he finds time; there is no particular hurry.