[Lucia, Terrence]

Similar to 14800 (Now it starts to get inconsistent, sorry). This time we were unhappy with the LVDT diagonalization so we proceed to make it better. This time we also tried something new and optimized the complementary filter differently. To see if it is possible to lower the blending frequency and enable active isolation at microseism, during the optimization I have used weighting functions to amplify the noise between 0.2 and 1 Hz by 10 times. The resultant complementary low-pass have a notch-like feature that quickly rolls off, thus increasing the contribution from the geophone. The result is a complementary filter that blends at 150 mHz which is the lowest amongst all other suspensions for now. As a trade-off, the noise a lower frequencies are also amplified and is higher compared to other suspensions. The inertial loop was closed without any problem. We should consider whether or not such tradeoffs is necessary for enabling active isolation. For now I don't have an answer to that. Later I will have yet another approach to synthesize complementary filters, we will see if the new method is better before doing any changes.