[Park, Cho, Kokeyama]
Tiltsensor is reinstalled at the MCE pylons. The oplev transmitter and receiver were moved to accomodate both of the sensors. Photos and details will be posted later.
[Park, Cho, Kokeyama, 1/22-26]
Here are photos of the installed tiltsensor at MCE. Both oplev and tiltsensor are installed together in the set of the pylons for dedundancy. As the windshields, just tin foils are covered over the pylons.
There are two spectra attached when MCE PIT or YAW was excited, respectively. Both oplev and tiltsensor calibrations seems OK. As expected, the tiltsensor has better sensitivity above 10 Hz where the oplev signal is limited by the electric noise. In lower frequencies, we don't know why the tiltsensor has higher noise floor.
The tiltsensor's real time models are running on k1imc0. k1pemimc0 (used for the workshop) was killed as it's not used anymore. The original tiltsensor model have problems where arc tangent cannot be used twice in one model, so two models (k1vistsp and k1vistsy) are running for caliculating PIT and YAW signals separately. We'll keep try to solve this issue.
The real time software issue was solved, thanks to Yamamoto-kun and the CDS test bench!
We found that there was VERY strange technical rules when two arc-tangents are used in one model. The first the model needs OSC (stragne!), and the second point is that any filter banks cannot be used in the same layer of the arc-tangent part (cdsFunctionCall). Very strange.
Anyway, now two arc-tangents run in one model. So only one model; k1vists is running on k1imc0 machine now.