Yesterday I did a DHARD_Y noise injection that had band-limited noise between 1.5 and 3.5 Hz. In the first plot I compare ASC-DHARD_Y and LSC-DARM in quiet times and during the injection.
There are several interestiing observations:
- DARM see an increase of noise at the excitation frequency and at the doubled frequency (and maybe other harmonics too)
- DARM noise between 10 and 80 Hz is also increased during the injection, even though DHARD_Y was not more noisy in that band
- there is little coherence between DHARD_Y and DARM, even when injecting noise
- the coherence between DHARD_Y and DARM at 20-60 Hz iss actually lower while injecting noise
This is good evidence that DHARD_Y is causing some non-linear noise coupling in DARM. This is also visible in the bicoherence of DARM with itself or of DHARD_Y and DARM.
It would be interesting to repeat similar noise injections, but with shaped noise and amplitudes such to increase only slighlty the DHARD_Y signal: in this way we might be able to estimate the efffect on DARM in normal conditions.
Given this finding, I was curious to see how CHARD_Y compared between H1 and L1. Attached are comparison spectra from an Observe mode lock a couple of days ago. There's nothing really suspiciously different between the two. There's a 1.5 Hz peak prominent in H1 but not L1, and the L1 noise above 15 Hz is much lower due to the more aggressive low-pass filtering.
