Continuing work from alog 65220, today RyanS, Naoki, and I went into the PSL to see if slightly changing the alignment of the first EOM in the PSL path (which sits near the NPRO, and is one of the actuators for the FSS) affected the 2.5 MHz peak that the squeezer team found.

We carefully moved each of the 4 adjustment screws on the EOM mount, but none of them seemed to have any effect on the peak (our Agilent was again hooked up to the TTFSS box's IN2).  Since I didn't want to move so much that I would affect alignment to the amplifiers, I kept my knob-turning small enough that we only barely could see a change in power transmitted through either amplifier.  I think the most I turned any of the knobs was 1/4 turn.  This affected the output of amplifier 1 by something less than 0.1 W (out of 70 W). 

Interestingly, there seemed to be a few extrema in the EOM alignment that brought the 1st amplifier's power down by a tiny bit, but at the same time brought the 2nd amplifier's power up by a tiny bit.  I wonder if this means that the amplifiers could use some input alignment tweaks themselves?  We didn't explore this further today.

After coming out of the enclosure, Daniel back-of-the-envelope estimated that when the peak is present, it could be of order 1 percent of the light going in to the PMC, so of order 1 W of light lost to this mode.  That seems not so great.  Daniel points out that we could consider trying to get a more direct measurement by looking at the beam at IMC REFL.  For that, we'd have to see what the response of the IMC REFL PD is at 2.5 MHz, or we'd have to put a fast photodiode temporarily somewhere in a low power path like the camera path.

While one could think about notching this peak on the squeezer side of things, it obviously seems better to mitigate this peak at the source.  We think it may take some thinking to add a notch to the generation of TTFSS boxes that we're using in the PSL, so for now we've left the FSS at the lower Common gain of 11 dB, so the bandwidth of the loop is around 230 kHz, but the 2.5 MHz peak is not there. 

Jason noted to me this afternoon that in addition to the 1 generation newer (but still pretty old) TTFSS box that is in the optics lab FSS test setup, we have a spare of the current generation TTFSS box (either in the optics lab or in the PSL anteroom).  We should have a look at that box, ensure that it has the 1.8 MHz notch for the lower mode of this EOM (which our in-use TTFSS has), and see if we can add a second notch even though there isn't a prepared space for one in these older generation boxes.  Since this will take a bit of time, for now we're going to just see if we can work with the lower bandwidth FSS for a while.