Reports 1-1 of 1 Clear search Modify search
VIS (EX)
yoshinori.fujii - 14:41 Monday 12 August 2019 (9941) Print this report
A sensor correction test with IP

K. Miyo and Y. Fujii,

log on August 9th 2019,

= ETMX =

  • We have conducted a sensor correction test mainly focusing on a bump at around 140mHz, which have been observed at all the stages, i.e., IP, BF and TM so far.
  • We could suppress the TM-L amplitude at 150mHz successfully by a factor of about 2.7.
  • This was true (at least) when the seismic level was high.

== Setting and what we did ==

  • The IP was controlled by only LVDTs since the inertial damping with TMMA-accelerometers (for ITMX), with geophones (for ETMs) were still under characterization.
  • We subtracted seismic signal from IP-LVDTs;
    • we added local seismometer signal (in displacement) with -1 and with 3mHz cut-off to the IP-LVDT sensor signal. when the cut-off frequency was set to lower frequency like 1mHz, the IP started to oscillate at low frequency and it made the TM-L amplitude larger.
    • This subtraction was done only for IP-L dof (just for a test).
  • Figure 1 shows the time series of the IP-L and TM-L w/ and w/o the subtraction.
  • Figure 2 shows displacement transfer functions from the ground motion in L,  w/ and w/o the subtraction;
    • red-dashed: IP-L signal sensed by geophones (inertial sensors on IP) w/o the sensor correction in [(um/s)/um].
    • magenta-solid: IP-L signal sensed by geophones (inertial sensors on IP) w/o the sensor correction in [(um/s)/um].
    • blue-dashed: TM-L signal sensed by length-oplev w/o the sensor correction in [um/um].
    • cyan-solid: TM-L signal sensed by length-oplev w/ the sensor correction in [um/um].
  • This did NOT led larger TM oscillation in P and Y as Figure 3 shows.
  • The measurement files are stored in: /users/VIS/typeA/ETMX/LogNotes/20190809/
  • To be continued

== For next ==

  • implement the correction filter to IP-T and see the vibration isolation ratio to the TM-L.
  • check this sensor correction effect to the TM when the seismic noise level is low.
  • compare the vibration isolation ratio with the following filters:
    • this sensor correstion at IP
    • band-pass filter at BF (like ITMX, klog#9936)
    • (usual filter at BF (like ITMY) )
Images attached to this report
Comments to this report:
ryutaro.takahashi - 9:50 Tuesday 13 August 2019 (9946) Print this report

This is a good job.

I understood that this sensor correction is a kind of feed forward control. If so, I wonder at the oscilation of IP in the lower cut-off frequency.

yoshinori.fujii - 13:34 Wednesday 14 August 2019 (9965) Print this report

This filter is the one which subtracts the seismic motion signal from IP-LVDTs.

> If so, I wonder at the oscilation of IP in the lower cut-off frequency.
Yes, we found lower frequency oscillation when the cut-out frequency was set at a few mHz. Now it is at 30mHz, the oscillation became smaller conpared to a few mHz cut-off case. We might still have to check/tune the cut-off frequency.

Search Help
×

Warning

×