Reports until 16:52, Thursday 22 September 2022
H1 TCS
ross.johnston@LIGO.ORG - posted 16:52, Thursday 22 September 2022 - last comment - 17:10, Tuesday 11 October 2022(65075)
ETM HWS Ring heater test
Manually set the Ring heaters on both the ETM X&Y to run at 2Ws overnight. Will return in the morning to see if we can discern any shifts in the dots: compared to today's reference image.
Comments related to this report
ross.johnston@LIGO.ORG - 15:12, Monday 26 September 2022 (65113)


		
		
Images attached to this comment
camilla.compton@LIGO.ORG - 15:07, Tuesday 04 October 2022 (65195)

Comparing both ETM HWS to their simulation models during power up and Ross's alog65075 ring heater test.

The shapes look okay for the ring heater tests. EY has a strange bump when the RH was turned off.

Can't see a clear shape in the IFO power up but both spherical powers do increase as expected.  ETMX is clearer. Looking at the simulation numbers, we except a factor of 4 less in spherical power change during a 50W power up than RH test. This should be seen but is only a factor of 2 or 3 above the noise. We should work on reducing the noise and having the total power on the HWS CCD constant during power ups. 

  Power up: 2W to 50W PSL input (Noise level) RH Test: 2W/seg to 0W (Noise level)
  Simulation (x10-6 diopters) Measured Spherical Power (x10-6) Simulation (x10-6 diopters) Measured Spherical Power  (x10-6)
ETMY  + 17  + 700 (+/- 400) * + 66 + 150 (+/- 15)
ETMX  + 18  + 290  (+/- 85) + 66 + 1050 (+/- 80)

 * The levels are different as we changed the camera frame rate settings - wiki on camera

It used to be thought that the EY HWS spherical power was opposite to expected when powering up (alog 52839, 62340), but I think that we could have been just getting confused by non-physical spherical power changes when the total power on the HWS CCD changed (H1:TCS-ETMY_HWS_PROBE_TOTAL_PIXEL_VALUE). See attached 2019 spherical power data at time of alog 52839
Aidan saw this in 62098
Images attached to this comment
camilla.compton@LIGO.ORG - 15:46, Tuesday 04 October 2022 (65216)

It seems that we can blame the ITMY retro-refelction for the ETMY HWS total power and therefore spherical power change. See attached image with HWS channels coresponding with ITMY oplevs for 2019 data and 2022 data.

This is not as coupled in ETMX,see attached.

Images attached to this comment
camilla.compton@LIGO.ORG - 17:08, Monday 10 October 2022 (65286)

The ETMY power-up data above isn't correct as I chose a time that the Hartmann plate was off (alog64385). Updated power up data on 12th Aug 2:00 UTC attached. Updated chart below, see ETMY noise >> measured change!

Sheila also pointed out that the total power on the ccd changes may not be from the ITMs as both the ETMs and ITMs alignment change at the same time when we power up. 

  Power up: 2W to 50W PSL input (Noise level) RH Test: 2W/seg to 0W (Noise level)
  Simulation (x10-6 diopters) Measured Spherical Power (x10-6) Simulation (x10-6 diopters) Measured Spherical Power  (x10-6)
ETMY  + 17  + 10 (+/- 20)  + 66 + 150 (+/- 15)
ETMX  + 18  + 290  (+/- 85) + 66 + 1050 (+/- 80)
Images attached to this comment
ross.johnston@LIGO.ORG - 17:10, Tuesday 11 October 2022 (65303)
It is clear from today's deliberate misalignments shifts, equivalent to those that were produced during a previous power up and as shown in Camilla's last scope, that today we see an approximately a 10% drop in probe power incident on the HWS ETMY CCD. Although the absolute change is similar the relative drop in power is not the 20% that was shown in the above scope. We plan to check if the ITM, which also moves during power up, independent contribution to the ETM's CCD power drop. 

I have also attached a repeat scan where the pitch and yaw were offset by roughly the same amount. The change of both appears to have equal contributions to the power drop and I wonder if a more detailed 3D plot of the power or spot count (z) vs both pitch (x) and yaw (y) would allow one to better isolate the clipping. In a simple sense, if the power drops before reaching the CCD's own clipping edge then you know that it has to be from a point further back in the path. Further investigations of this will follow shortly. 
Images attached to this comment