J. Kissel, J. Oberling, J. Driggers, G. Mansell, N. Lecoeuche, S. Dwyer Once we hit Nominal Low Noise (relatively painless process, no troubles like we've been having for the past ~week), we found that DELTA L EXTERNAL was severely glitch-y, obvious both on wall ASD and in the DMT OMEGA glich-spectrogram on the wall. In the DELTAL spectrum, glitching was occuring at high frequency, so we zeroed in on the major activity of the the day that may involve that part of the spectrum -- FSS work, Balun Replacement, and Squeezer changes. - We ruled out the squeezer quickly by closing the SQZ beam diverter. - We played around with the FSS common gain, and found that gliching got much worse with the common gain (H1:PSL-FSS_COMMON_GAIN) at 16 dB, vs. what the PSL crew had set it at; 19 dB. - Sheila realized the glitching looked like what happens when the IMC VCO is in a bad place for RF whistles. - We yanked around the IMC VCO frequency (unintentionally) to some quite different place, and found that all glitching disappeared. The above process took us about 1.5 hours to figure out. We're not *really* sure what has happened, but the message is now that -- in order to have no RF whistles -- the IMC VCO requires a quite different absolute frequency (78.5 vs. 78.8 MHz). The suspicion is that -- after the FSS touch-ups today, the RefCav is now locked on a 00 mode at a different FSR. We are now running the sweep through the IMC VCO's frequency (while in observe) to find the best region, and will report more this later.
Adding spectrograms of the OMC_PI channel before (top) and after (bottom) the IMC_VCO frequency was changed. This shows the whistles very clearly.
We went to Observing for a few minutes while the IMC VCO frequency was slowly sweeping. This is something that we've done in the past to try to find better whistle-free zones (ex. alog 47987).
After talking to Keita, since this sweep would definitely take us through some whistle regions, while trying to find a nice wide whistle-free space, I have stopped the sweep and am leaving the VCO TuneOffset at the -4.95 V place that we see is pretty quiet.
Since the IMC frequency hadn't changed, and the TuneOfs hadn't changed from the previous nominal of -2.10 V, something else must have changed to make our previously nominal IMC frequency be very whistle-y. So, we should figure out what else might have changed, or we should do a full IMC VCO sweep to see if we can find a quiet region that is nice and wide, but also a little farther from the edge of the VCO's range (we're quite far from the actual edge, but the VCO becomes less-than-linear as you approach the edge).
So, for the first few minutes of this Observe segment the IMC frequency was changing slowly a small amount.
Looking at things that might have changed today, I've attached a 1-day and a 3-day trend of the PSL NPRO crystal temperature; for those not familiar, this temperature is the slow control for the FSS and is what we change if we need to lock the RefCav on a different FSR. As can be seen, we're still locking the FSS RefCav in the same region as before this morning's FSS tune-up, but it does appear to be making a move higher over the last ~4 hours (the "flat-line followed by fast changes followed by another flat-line" on the rights side of each plot are our adjustments from this morning).
Taking the trend out to June 2018 (3rd attachment, the farthest back I can get ndscope to go at the moment), it is seen that we used to lock the RefCav with an NPRO temp of around 0.22 (we had moved up by one FSR to lock in the mid-0.3 area to make ALS happier after the 70W amp installation). It is also apparent that the FSS has run with a NPRO temp in the upper 0.3 area in the past, so the slightly higher move after we locked the RefCav this morning isn't entirely unusual.
