This morning I retuned the IMC alignment WFS offsets to reduce the coupling of jitter into the ISS array.

The results

I could reduce to basically zero the coherence between the PSL periscope accelerometer and the ISS second loop signal, with the IMC locked and ISS second loop open. The first plot shows that the periscope noise between 100 and 700 Hz reduced down to the background noise of the signal. The two lines at 101 and 153 Hz are respectively a PZT pitch and yaw lines I was injecting.

The new offsets are

The procedure

Here is the procedure I used to tune the offsets, in details:

1. Inject a 101 Hz line in IMC-PZT_PIT_EXC with amplitude of 3 and a 153 Hz line in IMC-PZT_YAW_EXC with an amplitude of 3
2. I wrote three python scripts to monitor the jitter coupling (attached to this elog). The scripts demodulate_iss_pit.py and demodulate_iss_yaw.py gives you a real time plot of the demodulated amplitude of the excitations lines in the ISS power signal. Each plot contains the real and imaginary part of the ransfer function at the line. Those traces turned out to be a wonderful signal to tune the offsets. See in the second plot how the traces go to zero when the offsets are tuned. The third script online_blrms.py gives you another real time trace of the band-limited RMS of the ISS signal in the 200 to 400 Hz, thus measuring the noise due to the jitter peaks. Notice in the second plot how the noise reduced a lot when the yaw offset was tuned.
3. You can run the three scripts from any terminal, at the same time, and watch the traces as you change the offsets
4. Today the yaw offset was the worst one, so I tuned it first. here is some cross coupling between yaw and pitch, so chancing the yaw offset had an effect in pitch too. With a few iterations I converged to the final values above.