Last week we continued investigating the noise in RFC error signal, particularly at frequencies that we found coherent with DARM (41824, we show this again in Figure 1).

Since the cure of the 107Hz noise (last week, 42212) the RFC signal is cleaner and coherences between RFC and accelerometers around IB (IB tower, Crio Trap, external injection bench) are now more evident (Figure 2).

Besides the 101Hz peak (that we know from some time, 41948) there are very coherent peaks at 92Hz, 98.7Hz and 99Hz. The 92 Hz peak might be again the turbo pump fan (38978)??

The 101Hz peak has no evident similarly shaped peak in accelerometers (Figure 3) but the 99, 98.7Hz and 92Hz ones do (Figure 3 and Figure 4).

The accelerometer that sees the by far the largest coherence at 101Hz is the accelerometer on the North flange of IB tower facing the cryotrap and PR (named ENV_IB_ACC_X).  Although we found the 101Hz could be easily excited tapping on IB EAST flange (41947) it seems worth to explore tapping on the IB North flange as well.

The 92 an 99Hz peaks have the largest amplitude in the IB_ACC_X accelerometer.

In order to check if these peaks are real seism or instead some fake electrical noise polluting the accelerometers, Federico did this:

1) detached the IB_ACC_X accelerometer from the tower flange and seismically isolated it placeing it onto some foam and rubber (Figure 5, brownish curve)

2) add an electrical connection to ground so this accelerometer senses essentially ground currents (Figure 5, purple curve) 

The peaks at 92 and 99Hz disappear from the accelerometer in both configurations, proving they are real seism (zoom in Figure 6). Instead the peak in RFC at 98.7Hz seems to be ground current noise (Figure 6 as well).

What instead appears is a lot of coherence with RFC signal when in configuration 2). So it seems the RFC signal is largely polluted by ground current noise. This has to be further investigated.