JimW HughR
The first attachment is from late afternoon yesterday, before the winds started picking up. There is still a large difference between it and the ISI GND STS below 100mHz. This may be thermal equilibrating or maybe the masses need centering again. I'll look at this today; must be done at EndY.
The second attachment is from early this morning with sustained 20mph winds; pretty steady direction from the SW to SSW. The coherence and spectra look much better with the greater input.
The third attachment looks at coherence to the BRS Tilt. The coherence to the right is when the wind is low and to the left shows when the wind is high. Bottom line: not much difference between the colocated sensor and the GND seismo on the floor.
At the end of the commisioning break today, I went down to EY and recentered the masses on the T240. This sensor doesn't have the full integration into the data system, so it's not just a button push. Richards DB9 connector makes this pretty easy though. Pin 5 is the analog ground, so UVW are read by read getting the voltages from pins 1-3 to 5. To re-center, you need a 5V supply (Richard loaned me one), connect pin 7 to the supply ground and put the +5v on pin 6 for a(n actual) second. When I got to EY the masses were at 1-2v, after centering they were between .08-.2v.
I looked at 2-k seconds of data from Wednesday night at 6:00 UTC. Plots are attached. The ASD is in rad/rt(Hz). Once again, the T240 seismometer on the platform sees a substantially larger low-frequency signal compared to the BRS or the ground STS. This instrument performed much better when it was on the ground.
This suggests that it is picking up extra signal when it is on the platform, which is not real tilt (as the BRS signal is much smaller). It is unlikely to be temperature since Hugh's additional foam insulation made no difference. It could be temperature noise on the table feet so it may be a good idea to wrap the feet also in the same foam insulation and see if that makes any difference.
Another possibility is that the low-frequency excess could be some form of (parametric) down-conversion from the excess high frequency motion of the platform (10-50 Hz). We have seen such noise on the BRS flexures, so it would not surprise me if similar noise exists in seismometers. I'm more surprised that the platform is shaking so much at high frequencies. I think this calls for a better design of the platform feet.
