SR3 oplev was realigned to the new SR3 angle.
We've found that this oplev had a clipping problem even before SR3 angle was changed. OPLEV SUM used to be about 30000 counts, and after we were done the SUM jumped up to about 66000 counts without changing the whitening setting. Doug thinks that this might be a bit too high a count, but ADC is not railing so we'll leave it as is.
Details:
SR3 oplev had two serious problems that are not related to SR3 angle.
- There was a large clipping inside the beam launcher enclosure (first picture). This was fixed by moving the launcher inside the enclosure.
- The receiver enclosure and the viewport are connected by a metal pipe, but the bellows at the end of the pipe was not strong enough and the thing was sagging badly, making the aperture much smaller than it should be (second picture). This was temporarily fixed by putting some material under the pipe (third picture). This is not a real fix, I think.
After these were fixed, Doug aligned the launcher such that the return beam comes through the receiver viewport/pipe cleanly. Due to the beam angle shooting up from HAM5, in order to avoid clipping by the receiver viewport, bellows and the pipe, the beam was positioned much higher than the center of the pipe opening at the receiver box.
At this point, we recognized that the oplev sensor couldn't be moved high enough to receive the beam as the vertical stage doesn't have enough range. Actually the vertical stage has a huge range, but at least 3/4 of that range is wasted to be able to lower the sensor to the position where there is no possibility of receiving any beam with any SR3 alignment.
We removed the aluminum part on which the sensor is mounted, Doug made new holes, we remounted the sensor at a higher position (4th picture) and the translation stages were moved until the beam was roughly centered (5th picture), and we used the MEDM screen to fine tune.
A cell phone camera with selfie mode turned out to be quite useful to monitor the beam position when there's no light on the sensor, and that's what was used to take the 5th picture.
The whitening board is already on the lowest gain setting, so I think we cannot reduce the counts that way.
I have turned off the single whitening stage that was on previously (LHO#14756). This won't help if the optic makes large excusions in pitch/yaw and saturates the ADC, but we are so far above the ADC noise that I think whitening is not necessary here. (See first attachment. Blue is with whitening, red is without whitening.)
The SR3 pitch oplev damping loop appears to work fine with no modifications. (See second attachment. Blue is loop off, red is loop on.)
