Online Data Analysis (Omega)verkindt - 12:16 Monday 29 August 2011 (30196)
Print this reportinvestigations on omega triggers between 300 and 500 HzDuring discussions around glitch classification,
Florent mentioned a "line of glitches" around 460 Hz whose source may be a non-stationary frequency line around 460 Hz, described by Gabriele as "first order violin modes of the last
stage of cavity mirror suspension". In June, such, there are lots of such 460 Hz glitches.
Looking at today's omegagram, we can see a lot of loud (snr>10) glitches also around an other frequency near 500 Hz (plot 1). The corresponding omega triggers are tagged with a frequency 498 Hz or 493 Hz. Plot2 shows the dark fringe spectrogram, early today. We clearly see some lines (some of them non-stationary on a short time scale) but none are around 500 Hz.
Computing band rms around each frequency line visible on the plot 2 and looking at the spectrum of it (plots 3 and 4), I found the usual 0.6 Hz for several of them but also some unusual (to me) low frequencies:
spectrum peak for band rms 350-352 Hz : 0.5 Hz
spectrum peak for band rms 443-445 Hz : 0.18 Hz, 0.67 Hz
spectrum peak for band rms 463-465 Hz : 0.07, 0.14, 0.17, 0.22
spectrum peak for band rms 468-470 Hz : 0.07, 0.14, 0.17, 0.22
But no hint for the 493 Hz and 498 Hz glitches. It would be interesting to know better how the frequencies of the omega triggers can be correlated to frequencies of non-stationary lines.
It would be interesting also to have some hints for the non-stationarity low frequencies above.
Images attached to this report
Comments to this report:
Vajente - 13:36 Monday 29 August 2011 (30198)
Print this reportThe region between 457 and 464 Hz is crowded by a large number of violin modes (very narrow spectral lines due to the resonance of the monolithic fibers suspending the arm cavity mirrors. 4 mirrors times 4 fibers per mirror makes at least 16 lines). Since mirrors and fibers are very similar one to the other, these lines are very close in frequency. The variation of the BRMS computed over a 2 Hz wide band is expected to contain all the beating frequencies between the violin modes.