[Okutomi with Miyo-kun's help]

I characterized three geophones for Type-A ETMX control test.

The geophones have better noise level than the expected accelerometers!

Target

To characterize the assembled geophones with measurements of

• Efficiency: a conversion from raw outputs to data in velocity or displacement
• Noise: sensor noise in frequency domain

Issue

• The geophone L-4C has each inspection sheet which gives the information of the conversion efficiency
• But their individual parameters seems not to be valid any more
  • Because a simultaneous measurement with the geophones and PEM seismometers has different level of ground motion to each other using the parameters on the inspection sheet
• So I estimated the parameters of geophone with a free run measurement to get the conversion efficiency 
• Then the converted sensor noise of the geophones is obtained

Free-run measurement

• Geophones
  • In-pod x3: with pre-amp circuit, set on each aluminum plate on the floor (for tilt adjustment)
  • Naked: without pre-amp circuit, set on a table 
• Seismometers
  • Trillium Compact x2 (JGW-S1707526): 45dB amplification with whitening filter, set in a corner of EXV room (outside of the clean booth)
  • Trillium 120QA (JGW-T1707554): 45dB amplification with whitening filter, set in a corner of EXV room (outside of the clean booth)

• Geophone parameter
  • Generator constant [V/(m/s)]: assumed that the values in the inspection sheet are correct
  • Damping ratio: to be estimated
  • Resonant frequency (f0) [Hz]: to be estimated
• Geophones' parameters are adjusted so that the geophones and the seismometers have same spectra of ground motion in the range of 0.1-10 Hz
• In this parameter estimation, the seismometers are regarded as reference because of
  • broadband flat response (0.01-10 Hz)
  • the fact that the 3 seismometers have a consistent level of displacement spectrum even with a simple calibration method

• Estimated geophone parameters

• Notes:
  • The pre-amp circuit (JGW-D1301466) seems to have a gain of 375.53 * 5.0 / 2 = 938.8, not 375.53 * 5.0 = 1877.7
    • Sekiguchi-san's PhD thesis (p.114 in JGW-P1504155) missed to mention that the differential opamp at the final stage outputs a half amplitude of the input voltage
    • You can check that in the data sheet of the differential opamp => product page
  • Trillium Compact: 749.1 [V/(m/s)] (measured, cf. JGW-D1707542)
  • Trillium 120QA: 1202.1 [V/(m/s)] (nominal, cf. JGW-T1707554)
  • The disagreement between the geophones and the seismometers in heigher frequency (> 10 Hz)would come from the accuracy of the pod assembling, the base plate where the geophones are put and so on
    • This disagreement may not be a problem because the frequency band above 10 Hz is out of the range for the geophone to monitor velocity of the IP stage

Noise measurement

• Sensor noises are measured using the conversion efficiency obtained above
  • Putting the geophone pod upright makes the internal proof mass fixed

• Geophone sensor noise

• Accelerometers which we "were" going to use for Type-A are expected to have a noise level of ~ 3 x 10^-8 m/Hz^1/2 @ 0.1 Hz (according to a measurement in TAMA?)
• The geophones now we have can achieve smaller noise level than the expected accelerometers

Current situation

• Now the geophone pods are installed on the IP of Type-A ETMX
• The response we obtained needs to be implemented into the input filter on the digital system
• Sensor configuration on the ETMX IP: