Evan, Elenna
Evan took a look at many of our ASC cutoffs today while we investigated the 2 Hz yaw instability. Many of the cutoff filters we have in place are unnecessarily stealing phase in the loops. They might not be our actual problem with the ASC loops, but it was a good fix to apply. Evan adjusted the Qs in the elliptical low pass filters and was able to reduce the phase loss of many of the filters. They have been saved and tested, and added to the guardian.
DSOFT ad CSOFT Y: these loops are now cut off at 8 Hz, which was a change I had applied to reduce the soft yaw contribution to the noise budget. The new filters are in FM5 for both loops
CSOFT P: this loop had two different cutoffs at 10 and 13 Hz giving 60 dB of attentuation, which seems high. He maintained that attentuation and combined them into one 10Hz cutoff filter in FM 8. I think we should investigate if we really need 60 dB of attentuation in this loop.
CHARD Y: we removed the "newcutoff" filter I applied with the loop change, and adjusted the "cutoff18" filter. Now in FM8
CHARD P: new filter in FM6
Today I also fixed the DSOFT P cutoff filter. It was in the same configuration at CSOFT P where it had two cutoff filters, one at 10 Hz and one at 13 Hz. I have copied over the single cutoff for CSOFT P into FM10 (overwriting the ELP13 that was there). Again, I am not sure 60dB of attentuation is necessary for DSOFT or CSOFT P. Worth investigation.
Evan has determined that there is no improvement to be made with the DHARD P and Y ELP20 filters. However, I have moved some DHARD P filters around because the ELP20 was in FM1 and the control filter was in FM10 (divide by 1000, multiply by a thousand, etc.). I moved "ctrlHBW" to the FM6 place, ELP20 to FM10, and the 30 Hz notch that was in FM6 (currently unused) into FM1.
While unlocked, I turned on the proper filters for the ASC loops and SDFed the change. This doesn't really matter since these are all guardian controlled, but it seems like a good idea because then we can see the actual filters engaged in full lock all the time.
I have also updated the guardian to ensure the correct filters turn on at the right times.
I also have made plots of all the new filters, comparing them with the filters previously in use. The screenshots are attached (sorry for plot overload!). All screenshots show the old cutoff filter(s) in red, and the new cutoff filter in blue. Final screenshot is SDF, and I definitely hit confirm.