georgia.mansell@LIGO.ORG - posted 16:18, Wednesday 07 April 2021 - last comment - 16:17, Monday 26 April 2021(58522)
HWS work today
Camilla, Georgia
EX HWS
After yesterday's ring heater test we decided to touch up the alignment into the vacuum system. We used pico’s for fine alignment of HWS M1A.
We realised Hartmann plate was on backwards (the bolts attaching the front plate to the spacer where touching the camera, leaving a ~2mm gap between the hartmann plate and the camera), fixed it.
We measured 8.5 uW out of fiber collimator, and 2.9 uW after the beamsplitter. We estimate 20nW on Hartmann camera.
Camilla ran the HWS dead pixel's script according to Danny's instructions.
EY HWS
We found the light source on the table, and the power supply in the VEA, as left by Keita.
The beam is misaligned and clipping HWS M4.
We measured 0.2mW out of laser.
We didn't have time to work on the alignment, and plan to come back on Friday, unless prevented from doing so by commissioning.
We key'd the laser off and trasitioned the VEA back to laser safe.
Photos from EX showing the before and after alignment.
Old dark pixels using a threshold value of ~250 is shown in attached image. I had to work backwards to find out what threshold was used by comparing the generated dark pixel plot to the mask used in our ring heater test earlier in the week (see green dots on modified image). We moved the old dark images (from 2018-12-12 ?) to h1hwsex dark_images_old and the take_darks script automatically copies the new bad pixel numpy array to /data/H1/BAD_PIXELS/history/. I saved the array that made the attached image as ETMX_pre20210407.npy
We had some permissions errors when running the take_darks script located in tcs/common/scripts/take_darks which Dave has now fixed. New dark pixel image (attached) using same 250 threshold setting doesn't make sense. I think the problem could have been that we were most likely streaming the image of the camera at the same time as taking the dark images. If you try to stream the image twice it causes flashes to come across the screen so I think that's what we are picking up. We could try and take this again remotely by turning off the SLED and ALS rather than covering the camera.
New dark pixels HWS EX image attached. I checked that no image was being streamed before taking the dark images but we still have lines going across so I'm not sure what the problem could be.
Taken by running >>
bash /opt/rtcds/userapps/release/tcs/common/scripts/take_darks/take_darks.sh 100 ETMX 250(i.e. 100 images with a threshold of 250) from within h1hwsex with ALS shutters closed and EX HWS SLED OFF.The results from the EX HWS are very confusing with the bad bad pixel subtraction.
I had a look at the raw files generated by the bad pixel scripts Camilla describes above, thinking one of them would have glitch and I could delete that file. Turns out all the images I looked at have streaks through the image (see attached screenshot).
When ALS is shuttered next I'll try taking the dark images used to calculate the bad pixel array again.