I moved all three WEB magnetometers onto WE tower base (S side), preserving the same orientation. W and V measure larger magnetic field at all frequencies. Also non stationary line at about 8Hz. See no clipping, but not far (clipping levelis around 60nT, presently max is around 50nT)
Having the magnetometers closer to the mirror we can reveal some signals that were not well visible until now.
Moreover, there are structures in hrec (e.g. between 40 and 50Hz) that have a little but not trascurable coherence with these sensors.
A possible path could be that these magnetometers now can read the Zcorr signal sent to the mirror's coils, but their coherence with this last is near zero, so it seems the path is a real magnetic field directly acting on mirror's magnets.
Yesterday we had the chance to have the ITF in comissioning mode for a short period, and we injected magnetic noise in WEB using the usual setup, as described in this logbook entry.
We were able to generate a white noise a factor of ten above the background, as read by sensors nearby the mirror, and a fleble but not zero coherence is visible between sensors and Hrec.
Later we also injected a comb of lines (11Hz spacing), some of which are visible in Hrec
Further analysis will follow
The time of the injections is visible in the spectrogram (third plot)
Here a tentative measurement of transfer function of far-field magnetic fields at WEB using the injected line comb (June 6).
Figures 1,2,3 show the injected comb (N,W and V magnetometers), 11 Hz spaced, done issuing manually one ACL_RAMP_CH command in ENVnoise to generate a saw-tooth like signal.
Figure 4 shows the TF (calibrated hrec/nT). Only the first 3 lines (11,33,44Hz) produced a visile effect in Hrec, elsewere the TF is just an upper limit. Figures 5 and 6 show the TF for the other two magnetometers.
Figure 7 shows the projection. It falls at least a factor 10 below order of the sensitivity. Figures 8 and 9 show the projection for the other two magnetometers.This result is consistent with the white noise injection (see previous comment by Federico).
Conclusions:
- This result indicates that the WEB magnetically coherent noise in Hrec that we observe around 40-45Hz should not couple through the TM magnets.
- Measurements have to be repeated injecting more intense lines at high frequency to better measure TF (next Tuesday calibration will inject the usual set of lines at WEB, 46080)
- The sawtooth injection coulb be an alternative method to injecting a lines in sequence as currently done by the MagneticInjection process. However the gain is not probably much. Figure 10 shows another test sawtooth injection that we did on June 11. The current amplitude is set to the maximum (avoiding distotion and saturaton of theamplifier) and the amplitude of magnetic lines is of few nT, rapidly decreasing at high frequency (but we can gain as sqrt of injetion time) while with the current way of injecting we have at least 1nT at all frequencies (up to 500Hz). The two methods are likely equivalent.
