KAGRA Logbook

MIF (General)

takafumi.ushiba - 16:02 Sunday 27 November 2022 (23051)

Recovery of REFL light power

[Yamamoto, Ushiba]

Abstract:
We found the power on REFL PDA1 decreased around 1:45 JST today.
After several investigation, we found the IFO REFL HWP is not rotated when we moved it to REV direction from the nominal position.
We adjusted the power on REFL PDA1 tonominal values, and currently HWP angle is 55 degs.

Detail:
Since Yokozawa-san reported that FPMI cannot be locked (klog23050), I invstigated the reason.
Then, I found that laser power on REFL PDA1 was very small (0.001 mW).
We investigated when it happened and found that it was about 1:45 JST (fig1).
However, there are no large suspension angle changes at that time.

So, we entered the mine and check the REFL table and confirmed the beam is comming to the REFL.
Also, the beam shape on the sensor card looks good (seems no clipping).
Then, we checked the HWP angle was nominal value (95 degs).

After that, we checked the log of HWP angle and REFL power, and then found that the REFL power didn't change when it was rotated from 95 degs to 70 degs.
So, I moved HWP so that the REFL power becomes almost nominal (0.11 mW).
After that, we checked the HWP motion and found that HWP cannot be moved to REV direction from there (fig2).
One possible reason is that HWP motor reaches the edge of its range and doesn't move to REV direction.
So, we need to investigate it and if so, we need to move the motor manually and save the range.

Note:
I'm not sure who rotated HWP for what kind of purpose but please ckeck the corresponding PD output (in the case of IFO REFL, REFL PDA1 output) when moving HWP.

Images attached to this report

23051_1669531283_Screenshot from 2022-11-27 15-40-46.png

23051_1669531973_Screenshot from 2022-11-27 14-55-16.png

Comments to this report:

takahiro.yamamoto - 19:23 Sunday 27 November 2022 (23054)

A total degree seems to be incremented/decremented even if the HWP script has some kind of errors.

-----
A bash script (medm_hwp_calib.sh) increments/decrements only when the python code (agilis-p_control.py) returns '0' as follows.
[ $? -eq 0 ] && ${CAPUT} K1:SYS-HWP_${NODE}_STEP_TOTAL_DEG ${curr}

But current python code always returns '0' such as

if int(sl[sl.find('L')+1:]) != -340 and int(sr[sr.find('R')+1:]) != 340:
    print('Bad softwarelimit ',sl,sr)
    exit()

To do the proper error handling, we need to give a non-zero value such as

if int(sl[sl.find('L')+1:]) != -340 and int(sr[sr.find('R')+1:]) != 340:
    print('Bad softwarelimit ',sl,sr)
    exit(1)

And also, return code isn't set when some kind of error is caught up on try~except in each method of 'agilisp' class.
I'm not sure that program should be stopped when an error is raised or not.
But the final return code of python script must be non-zero value in order to skip to increment/decrement on the bash script for the MEDM interface.

satoru.ikeda - 15:24 Wednesday 30 November 2022 (23083)

Fixed to return values on exit when an error occurs from the corresponding HWP python script.

I also added Log output.
The folder is as follows: VIS name, but this was adapted from another motor-based script.
/kagra/Dropbox/Subsystem/VIS/Scripts/HWP/LogFiles/k1hwp{0,1,2...} _(iSerial).log
iSerial can be found at
HalfWavePlate

Non-image files attached to this comment

japanese.txt

satoru.ikeda - 10:41 Thursday 01 December 2022 (23089)

Changed log output destination

Old:/kagra/Dropbox/Subsystem/VIS/Scripts/HWP/LogFiles/k1hwp{0,1,2...} _(iSerial).log

New:/kagra/Dropbox/Subsystems/MIF/scripts/HWP/LogFiles/k1hwp{0,1,2...} _(iSerial).log