reshape BF-L/T filter?

[YoshinoriFujii]

Todo

• From https://github.com/YoshinoriFujii/Weekly-task/issues/24
• Basically for X-arm suspensions
• check the resG impact
• check if the following peaks can be damped with BF-LVDTs
  • [ ] 0.25Hz => probably impossible
  • [x] 0.4Hz
  • [x] 0.67Hz

[YoshinoriFujii]

On 2019.9.22 ITMY

• For ITMY, the gain was changed as any how since it seemed the feedback force looked relatively large:
  • [BF-L: from 1 to 0.5
  • BF-T from 1 to 0.25

[YoshinoriFujii]

ITMX BF-L

• start:

• already implemented(dashed) vs. w/o BF-L damping filter:

---

• already implemented(dashed) vs. BF-L damping filter (damp_sc)

• peak(or bump) at 0.4Hz was decreased though, peaks at ~0.56Hz and ~0.8Hz appeared. To be tuned again.

---

• already implemented(dashed) vs. BF-L damping filter (damp_sc_test)
• dmap_sc - notch0.4Hz - resG0.15

• Again, peak(or bump) at 0.4Hz was decreased though, peaks at ~0.56Hz and ~0.8Hz appeared. To be tuned.

---

• already implemented(dashed) vs. BF-L damping filter (damp_sc_test)
• dmap_sc - notch0.4Hz + resG0.4102

• looked improved at 0.4Hz. => - [ ] check the performance in long-term

[YoshinoriFujii]

BTW, peak at 0.25Hz

• At this frequency, we have

  • BF-L/T dip in the IP-damped-mechanical TF
  • dip in vertical (GAS) TFs:

• According to a measurement, it seems that the peak at 0.25Hz is from vertical motion

• since a the GAS filters have a dip, and thus damping with GAS filter seems difficutl.
• [ ] How about BFV?
• [ ] btw, how are the control loops look like? => https://github.com/YoshinoriFujii/Weekly-task/issues/27

[YoshinoriFujii]

ITMX BF T loop

• remove 0.4Hz notch

• the 0.4Hz peak was damped but is it shifted? ~see the long term~
  • yes.at 0.4 Hz peak, the pjhase shift is ~0deg. The phase should be compensated for velocity damping.

---

Then,

[YoshinoriFujii]

ETMX BFL test

• start:

• BF-: comp w/ and w/o daming at BF
• The cyan is BF-T motion though, so far, NF-L was moving as same level as BF-T.

[YoshinoriFujii]

ETMX BF-T

• just to remove 0.4Hz peak in the spectra.

• after modification

  • Not completely but improvements was there. I just stopped since this was not the crucial one now(it seemed.)

    Then

• go back "start" if something becomes strange.