Reports until 15:48, Thursday 19 July 2018
H1 ISC (ISC)
craig.cahillane@LIGO.ORG - posted 15:48, Thursday 19 July 2018 - last comment - 16:39, Friday 14 June 2019(42978)
ALS COMM PLL glitches caused by Y arm light pollution
We took a look at several ALS PLLs to check if the ALS DIFF/COMM glitches we saw last night were associated with them.
From Plot 1 we see that both the COMM and DIFF PLL control signals glitch strongly at the same time as the ETMY coils.

Plot 2: We ran a test where we locked ALS COMM normally and looked for glitches, which we found (red).  Then we shuttered the Y arm and the glitches in ALS_COMM went away.
Seems that the COMM PLL was partially polluted by Y arm light, which is close in frequency to the X arm light.

Solutions being considered are fixing the ISCT1 alignment/changing the polarization of the X vs Y transmission beam, or moving the VCO frequencies.

Attachment 3 shows the beat frequencies of the entire ALS system.  Seems that the X and Y arm beat frequencies are within 10s of kHz and the X and Y beatnotes could be getting confused on ALS COMM.
Images attached to this report
Comments related to this report
craig.cahillane@LIGO.ORG - 19:07, Thursday 19 July 2018 (42982)
Georgia, Keita, Craig

Georgia installed a 532 nm half wave plate at ISCT1 in the Y path while Keita and I installed a HWP at End Y to try and reject Y light from the ALS_COMM beatnote.  
We were unable to eliminate glitches in ALS COMM PLL CTRL.

Images attached to this comment
georgia.mansell@LIGO.ORG - 22:59, Thursday 19 July 2018 (42987)

Note: The half wave plate installed on ISCT1 is on the ALS-Y path, directly before the combining beamsplitter for the ALS-DIFF beatnote, i.e. between ALS-BS6 and ALS-BS7 on this handy layout of ISCT1.

After Keita and Craig installed the half waveplate at end-y, but before I installed the half waveplate on ISCT1, I noted that there was still a substantial beatnote on the DIFF PD. If all had gone to plan the X and Y path on ISCT1 should have been in orthogonal polarisations at this point. With the waveplate installed I could increase the beatnote pk-pk by a factor of 2.

Maybe somewhere along the line (steering optics, vacuum windows?) we are picking up some circular polarisation.

Investigations ongoing.
 

peter.fritschel@LIGO.ORG - 09:01, Friday 20 July 2018 (42989)

Just a note that LLO dealt with ALS-DIFF glitches by changing the beam interference on ISCT1, so that instead of interfering the ALS X beam with the ALS Y beam, each ALS arm beam is interfered with the PSL frequency-doubled beam. This is described in LLO log entry 28005.

jeffrey.kissel@LIGO.ORG - 09:34, Friday 20 July 2018 (42991)
Associated with FRS Ticket 11113.
sheila.dwyer@LIGO.ORG - 16:39, Friday 14 June 2019 (49925)

Note for future reference:  

This problem was caused by electrical cross coupling, not light.. 

https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=43002