40m QIL Cryo_Lab CTN SUS_Lab TCS_Lab OMC_Lab CRIME_Lab FEA ENG_Labs OptContFac Mariner WBEEShop
  ATF eLog  Not logged in ELOG logo
Message ID: 1165     Entry time: Fri Nov 12 12:56:47 2010
Author: Zach 
Type: Laser 
Category: GYRO 
Subject: time-varying RFAM 

 This morning, I started fresh with this error signal offset issue. I put a beam block in the cavity so that it would not resonate and tried the following things:

  • First, I just tried to null the offset by adjusting the angle of the HWP before the EOM (as I did yesterday).
    • I observed a drift in the error signal at roughly 1 mV per 2-3 minutes. This doesn't seem like much, but it is fatal to the lock loop.
  • Then I adjusted the input QWP to ensure that the beam was linearly polarized on the way in (by iteratively adjusting it and the HWP following it to make the power on one port of the PBS go to zero).
    • The drift was still there after re-nulling the error signal
  • Finally, I decided to analyze the beam coming out of the EOM by circumventing most of the experiment.
    • I removed the HWP following the EOM and installed two steering mirrors to direct the beam to the dedicated steering mirror of the REFL CCW PD (PDA225)
    • I then adjusted the power to a reasonable level and plugged the signal from the PD into the RF analyzer
    • Since pure PM should not show up in the spectrum (the carrier beats with the two sidebands cancel each other out), while AM does, I purified the modulation by eliminating the peak at fmod = 18 MHz
    • Within 5 minutes or so, the peak had grown out of the noise floor by ~20 dBm (noise floor = -98 dBm, peak height = -80 dBm)

Therefore, it appears that there is some (strong) time dependence to the amount of RFAM we get out of the EOM. It's not clear to me why this wasn't an issue before we put in the new PDs, but I guess several things have changed in the past week or two. The experimental setup is below.


ELOG V3.1.3-