Sheila, Daniel, Vicky

We're settling the OPO PZTs situation by deciding to feedback to OPO PZT 1, while DC-biasing OPO PZT2 using FCGS CMB slow out for now. OPO PZT 1 has a good TF, high UGF, but limited/low dynamic range, unlike OPO PZT 2 which has high dynamic range but limited bandwidth.

This causes a problem e.g. this morning, when green/IR co-resonance in the OPO occured around with an FSR around either ~5V or ~100V on PZT1, making the OPO lock nearly unworkable. To fix this, we cabled the FCGS CMB's slow out directly to PZT-2; note that this bias will NOT show up on any of our OPO-locking monitors, watch out for this. We will need to improve this temporary biasing solution down the line.

At first, Sheila and I were very successful in doing the opposite: with only 2mW green pump into fiber, we swept PZT-2 with various dc biases on PZT-1. This was successful in showing that biasing PZT-1 > ~50V can recover good mode-matching over the whole sweep range of PZT2, see these slow PZT-2 mode scans! This would've given us a high dynamic actuator and lots of range to ensure our PZT can access a lockable FSR. We tried locking the OPO like this with Daniel, but two issues: 1) too much gain that the CMB's common gain bottomed out at -32dB. Daniel lowered the cmb output gain by adding a voltage divider at 1/10 (w/221,2040 ohms), so we could lock PZT2 with -26dB common gain. Then we hit issue 2) unfortunately, the PZT 2 resonance being ~1.8 kHz limits the opo lock bandwidth to a few hundred hertz. We ended up setting OPO locking back to PZT1, with PZT2 having the dc offset bias to keep PZT1's co-resonant FSR's within the scan range. Data and transfer functions for PZT1 and PZT2 are attached.

Daniel and I also measured the CLF (2f, RF6) transfer function with these CLF CMB settings:  common gain = 25 dB, with 2 common boosts (0/10Hz, 20Hz/2000Hz) on. CLF measured TF is consistent with O3.