J. Oberling, R. Savage, E. Merilh

Noting some of the recent issues with RefCav relocking, we went into the PSL to take some measurements of the FSS to see if we could gather some clues as to what's going on.  We began by taking a transfer function and immediately noticed something odd, see first attachment (FSS2.tif).  Out around 775kHz there is a large bump that rises above the UFG (please recall that the UGF for the FSS is read at -10dB due to the design of the TTFSS electronics; in this case it is just below 500kHz), with a corresponding bump in the phase.  Looking at the last TF we took, the bump in magnitude was present, but smaller than it is now, with no corresponding bump in the phase; this therefore, is a change in the TF since mid-December.  Zooming in on the area (2nd attachment, FSS3.tif) shows it to be more than a simple bump, but a series of them culminating in a large hump at ~776kHz that rises almost 10dB above the UGF.  We took a spectrum of the IN1 port on the TTFSS (closest to the mixer) and while it looked a bit "spikey" (for lack of a better word), the area around the hump looked different; more flat, but with a dip and return (with a spike in the middle of this), followed by a small bump (see third attachment, FSS6.tif).  Thinking this could be an indication of an issue with the NPRO noise eater (the noise eater (NE) supresses the natural relaxation oscillation of the NPRO crystal, which tends to hang out in this area (each laser is different)), so we repeated the IN1 spectrum with the noise eater ON and OFF; see 4th and 5th attachments, FSS8.tif (NE OFF) & FSS9.tif (NE ON).  As can be seen, the relaxation oscillation peaks around 928kHz, with the hump starting above 800kHz and ending around 1MHz; with the NE ON, the relaxation oscillation disappears; in both, the area below 800kHz is unchanged,indicating that the NPRO NE is not the problem.

We're still digesting this info, but having a hump beyond the UGF that pops up above it is likely the underlying cause of our recent FSS instabilities; at this point we're not sure what's causing this.  Rick suspects that we may have some kind of parallel path resonance happening, or maybe there's a notch in this version of the TTFSS that we need to re-tune.  I've added 1 final picture, FSS13.tif.  In this pic the marker is at the current FSS UGF of ~486kHz; at 10dB/div, it is very clear that this feature at 776kHz is at least 10dB above the UGF.  Investigation continues.