Patrick, Terra (remotely), Kiwamu,

related log: 34660

During the commissioning window today, we did a test in which we actively excited PI mode 23 (ETMY, 32 kHz) to determine whether we can reproduce a same type of DARM noise which had coincided with high RMS of PI mode 23 yesterday.

Conclusions:

• When the rms of mode 23 became high yesterday (typically for a duration of 30 sec), DARM had a glitch almost at the same time, resulting in momentary broadband noise (LF - 100 Hz).  This happened multiple times.
• We confirmed that exciting mode 23 by the ETMY ESD didn't reproduce a similar type of noise or glitches.
• However, instead, we found that driving mode 23 caused narrow line noises at a few frequencies.

Message:

• We suggest not damping mode 23 unless the amplitude of mode 23 is too high for the interferometer to stay locked. (Note: mode 23 is not controlled by a guardian, and will remain off unless manually set)

[The test]

The results are shown in the first attached. It contains the following three different configurations for the damping of mode 23.

• No damping (yellow).
• Damping on with the nominal gain of +1000 (cyan).
• Damping on with an opposite sign in the control gain, -1000 (purple).

When the damping was on, it produced a few narrow lines (at 14.5, 64.0, 78.5 and 142.5 Hz). We are not sure why. As we flipped the control gain, it excited the mode and resulted in higher line heights. Looking at the control signal sent to the ESD, we made sure that the DAC was not saturating. By eyeballing the medm screen, the typical amplitude we sent for the nominal damping setting was about several 100 counts at the DAC.
 

[A hypothesis]

This is totally a speculation: as some interferometer variables vary (such as OMC alignment or something), it changes the coupling for mode 23 to OMC making a high readout value for mode 23. At the same time, this phenomenon causes a glitch in DARM. So in this hypothesis, mode 23 is just a witness of the phenomenon and not the cause.

The second attached shows the glitches that we had yesterday coinciding with high RMS of mode 23. The other PI modes which are also derived from the OMC DCPDs (modes 5, 12 and 22) didn't show any obvious events at the same time. So this OMC coupling hypothesis may not be trustable.