Personal tools
You are here: Home → Virgo → Data Analysis → Wide band events study
Navigation
Log in


Forgot your username or password?
 
Document Actions

Wide band events study

by Marie Anne Bizouard — last modified 2008-07-15 05:41 
Goal
  • Investigate if the origin of the wide band noise events observed in the burst analyses is due to laser frequency noise.
  • It has been observed in qscans that these wide band noise events are associated to low ferquency (2-10 Hz) glitches in some mirror angular control correction signals. It is especially true for Sc_BS_txCorr, but it's osbserved for other suspensions (NE, WE). A mechanism similar to the BoB events observed in C7 due to the misalignment of the cavities could explain these events.

Analysis

  1. During VSR1 a line at 1111Hz has been injected to measure the coupling of the laser frequency noise in Pr_B1_ACp. We computed the BRMS at 1111Hz +/- 5 Hz sampled at 20 Hz.
  2. We used the EGC burst analysis triggers. The frequency range is from 300 Hz up to 5 kHz.
  3. A coincidence window of +/- 100 ms has been applied to associate each EGC triggers to a BRMS value (the max inside the +/- 100ms window)
  4. We consider 2 events population: all events, events with ferquency band width > 3 kHz.

Results
SNR_vs_frequency.pngWide band triggers : frequency bandwidth>3 kHZ.

The central frequency of the highest SNR events is around 900, 2300 2500 and 3000 Hz.
SNR_vs_BRMS.pngCoincidence window to associate a BRMS value to each EGC trigger: +/- 100ms

The high SNR wide band events have in majorith a BRMS 1111Hz that lays in the core of the BRMS distribution.

A small fraction of these events are in the tail of the BRMS distribution.
KWsign_vs_SNR.png
Coincidence window: +/- 500ms
A majority of the high KW significance events are wide band high SNR events, but a good fraction of the high SNR events have a low KW significance.

Conclusions

The hypothesis that the wide band noise events could be due to laser ferquency noise is not proven!
It seems that some events could be associated to laser frequency noise, but the effect is marginal.
The use of the KW triggers on Sc_BS_txCorr does not seem to be very efficient to catch all the wide band noise. Furthremore there are too many high KW significance that are associated to low SNR events in the dark fringe. The use of BRMS might be more efficient.