Reports 1-1 of 1 Clear search Modify search
MIF (ASC)
dan.chen - 17:15 Wednesday 06 November 2024 (31516) Print this report
ASC control adjustment

With Ushiba, Tanaka

Summary

  • As reported on klog31507, the interferometer was not stable because of 1.1Hz oscillation, we tried reducing it by the adjustment of ASC control filters.
  • DHARD_P, DHARD_Y, CHARD_Y, DSOFT_Y filters were adjusted.
  • Then, the 1.1Hz oscillation on the control loops was reduced.

Details

  • We checked when the 1.1Hz oscillation appeared by changing the LSC_LOCK state and ASC_LOCK state and found the ASC loops seemed to make the oscillation.
  • So we adjusted the control loops of ASC.
  • OLTF measurement and filter adjustments for ASC
    • State: LSC_LOCK = PRFPMI_RF_LOCKED, ASC_LOCK = PRFPMI_ASC_ENGAGED
    • DHARD_P
      • Adjustment for K1_ASC_DHARD_P loop by using K1_ASC_DHARD_P_OLG.xml
      • Added -3.3dB to move UGF to 0.3Hz
    • DHARD_Y
      • Added -4.8dB to move UGF to 0.6Hz
    • CHARD_P
      • Because the suppression did not change, we did not change the filter.
    • CHARD_Y
      • Added -0.6dB
    • DSOFT_P, CSOFT_P
      • Because these UGFs are too low, we did not adjust them.
    • DSOFT_Y
      • Changed Q of 1.7Hz from 2 to 1.3
      • Added a notch(1,2,10)
    • CSOFT_Y
      • Looks fine! -> We did not change the filters.
    • The results are attached.
      • Red line is the adjusted result and blue line or green line is the measurement before adjustment.
      • The improvements on DHARD_Y, DSOFT_Y seem to be large.
Images attached to this report
Comments to this report:
kenta.tanaka - 17:59 Wednesday 06 November 2024 (31518) Print this report

After the adjustement, we measured the spectra of the ARM ASC error signals and Type-A oplev's signals. Fig. 1 shows the results. The dot lines are the signals before adjustment with PRFPMI_RF_LOCK, the solid lines are after the adjustement with PRFPMI_RF_LOCK. As you can see, the 1.1 Hz peak in the signals, especially ETMX and ITMX YAW oplev damped well.  It seems to be fine.

Images attached to this comment
Search Help
×

Warning

×