Jenne, Sheila, Kiwamu, Hang, Stefan
We tried to monitor the AS WFS gains during powerup. However, we didn't see any sign flips. Nevertheless, a combination of SOFT and SCR1 loops ran away anyway.
On the next attempt we again observe the onset of a 0.4Hz oscillation in pitch, which, depending on alignment, starts around 13W-15W, and grows until lock-break.
Jenne tried to move the soft offsets around to see whether there are alignments that are less suceptible to the 0.4Hz oscillation - not much luck.
Looking at the OPLEVs we noticed that mostly the xarm pitch is moving during power-up (but too slow for radiation pressure effects). Note that during this the soft loop error signals are kept zero, which suggests some non-linearity in the X QPDs.
One last attempt: We opened the SOFT loops, then went to 10W. We still get our "favorite" oscillation. We'll look at some lockloss plots when we're more awake.