Krishna

I measured the tilt transfer function (normalized to 1 at high frequencies) for Compact-BRS yesterday by driving it with a piezo stack located under it's platform. The drive would produce both rotation and vertical motion of the platform, but as d is expected to be small, only the rotational drive is relevant. The first pdf shows the response of the instrument compared to the drive at five frequencies - four to the left of the resonance and one to the right (used to normalize the TF). The resonance was at ~62 mHz and the measurement suggested a d = 100 (+/-10) microns. The error bars and limits are rough estimates and not exact. This sugggested an addition of ~3.8 grams to the upper mass to reduce d and obtain a final frequency of 47 mHz.

After a few attempts at trying to adjust the frequency (mainly due to inaccurate frequency measurements), I finally got it close to 49 mHz, which I deemed to be sufficient. The new transfer function measurement yielded the second plot, indicating d = 11 +/-2 microns. For this value of d, the translation rejection is given by M*d/I = 4.9 kg*11e-6m / (0.08 kg m^2) = 7e-4 rad/m. 

This is much better than needed for high frequency (>1 Hz) seismic tilt measurement, where seismic wa5velengths are short and consequently the tilt (in rad) is substantial in comparison to the translaiton (in m) . At 10 Hz, for example, the wavelength is 20 m, therefore a translation amplitude of 1 nm has an associated tilt of ~ 0.3 nrad. A more patient and careful tuning process can reduce d below +-1 micron.