Something we've been seeing and was shown in high wind condition by Irene is that SWEB moves a bit more than SNEB. We don't observe it in the bench signals, but those mix also the ground in, so are not to be trusted for this. See Fig1 for the first evidence: FFTs of both top stage blended signals. In Fig2, the TIME signals are shown.

Opening the loops doesn't change much, see Fig3. SWEB, despite having its IP mode at lower frequency, has less cut of high frequency motion. Interesting to see is the Q of this mode is about 30, whereas the SNEB mode (at 0.1 Hz) is about 150-200. One expects lower Q for lower f, but not this much. More a factor 1.4 squared in this case.

Then I injected in Z to see if there was any difference, see Fig4. No difference except in how much the IP was 'allowed' to move. In the TIME signals (Fig5) we see a factor of 10 or so! The noise signals were not written in the ffl with high freq so no transfer functions were made for this showing.

Checking the inloop LVDTs before blending don't show much difference, see Fig6.

Seeing the transfer function of SWEB with the local ground geophone gives a strange image: A flat line at 0.1 or so between 0.5 Hz and 5 Hz, see Fig7. Any comments? Perhaps we'll plan to open SWEB to inspect the IP stage when we're working on SDB2 during 17/18 May, possibly the IP stage is stuck/ rubbing on some cabling since we evacuated the SWEB MiniTower.

----

SPRB

The vertical actuator (yellow flagged over the weekend) was offloaded and the loops are closed again