I ran a series of jitter and frequency noise injections last Thursday to investigate how jitter noise couples. I ran the usual frequency noise injection while we were on one carm sensor. Then, I performed the jitter injections, but I ran a 4 minute injection instead of the usual 1 minute to get better resolution.

Using these injections, I calculated the frequency noise coupling and jitter noise coupling. Then, I calculated the jitter that couples through frequency noise by measuring the jitter-to-frequency coupling, and multiplying that with the frequency-to-darm coupling. I also calculated the coupling measured by all four IMC WFS DC channels (A/B, pit/yaw), where we usually measure the coupling from IMC WFS A for the noise budget. When I estimate the frequency noise, I do not apply any corrections to account for the fact that we usually have two sensors, so in some regions we may be overestimating the frequency noise by sqrt(2).

Frequency noise is about a factor of 10 below DARM, especially at high frequency. This has been the case since we returned from the OFI vent.

Jitter noise is well-measured up to about 2 kHz. I made a higher resolution plot zoomed in from 10-50 Hz to look at the coupling measurement of some of the peaks that Robert noted. I see strong coupling of two narrow peaks.

The jitter noise that couples to frequency noise to DARM seems mostly to be much lower than the direct jitter coupling, except for WFS B yaw at low frequency. Here is another plot that compares the direct jitter coupling, jitter-to-frequency coupling, and frequency coupling as witnessed by each jitter sensor.

I did a similar exercise with frequency and intensity noise in this alog.

The DARM trace shown here is actually GDS CLEAN, meaning that some jitter noise above 100 Hz has been subtracted in the trace. However, to calculate the coupling I used the NOLINES channel, which has no jitter cleaning.