While trying to scan the higher order modes in the OPO the green trans camera wouldn't show any beam. The camera turned out to be misaigned. This was fixed with a short incursion into SQZT6.
The plot below shows a typical scan. The next 3 snapshots show a TEM00, TEM01 and donut mode resonances, respectively.
The donut mode always seems to have an amplitude around 12-14% of the carrier, whereas the TEM01 mode seems to be larger at lower PZT values. The highest I observed was 25% (near 0V at the PZT) and the lowest around 10% (near 100V PZT offset). In the center of the range the TEM01 seems to be around 15%-20%.
I tried to verify this by locking at different carrier PZT offsets. With 23.3mW at the fiber launch point, when locked at -7V offset the transmitted power was at 89%, When locked at 49V or 92V, the transmitted power was both at 96%. So, at the very low end we seem to loose about 7% in the build-up power. Even so the trend is the same as with the scan results, it doesn't agree very well numerically.
Lookig at the trends in the last plot, we can see that the launch power as well as the OPO reflected green power has started to increase around mid August, while the transmitted power was kept constant. The reflected power has increased by about 50%, whereas the launch power increased by about 25%. This looks like the fiber has started to degrade again.
I check the actual volatge put out by the PZT driver. If you ask for 100V it will output 100V. However, the readback is about 5% short of this. Looking at the schematics, the monitor is a resistive divider 240K/12K that is directly fed to a Beckhoff terminal. From experience the later has about 250K input resistance.