ITMXのGASのループの確認

[YoshinoriFujii]

Background

• ITMXのBFのspectraに0.25Hzのピークがあった。

  • TML にも見えていた様子。
  • GASとhigh-coherenceあった。

• loopみてみたら理解できない形のものがあった。like:

• これ、どういう思想？

• これ、ここから入ってくるノイズが0.25Hzのソースの可能性あるのでは。

• ほかのsuspension 大丈夫？

[YoshinoriFujii]

status check

• Starting point:

---

• F1GAS

---

• F2GAS

---

• F3GAS

[YoshinoriFujii]

Modification plan

---

• F1GAS

• blue: the currently implemented one

• red: which would be tested, the plan (dcdamp_test) Maybe I want to have infinite at lower frequency region.

---

• F2GAS

• blue: the currently implemented one
• green: which would be tested, the plan 1 (dcdamp_test)
• red: the plan 2 (dc_test)

---

• F3GAS

• blue: the currently implemented one

• green: which would be tested, the plan 1 (dcdamp_test)

• red: the plan 2 (dc_test)

---

• BFGAS

• blue: the currently implemented one

• green: which would be tested, the plan 1 (dcdamp_test)

• red: the plan 2 (dc_test)

• [ ] BFGAS loop、dcだけなら閉じる？

[YoshinoriFujii]

2019.9.25 modification work

• starting point:

---

After modicfication

• Both "dcdamp_test" and dc_test" were working properly without exciting resonant modes (basically payload modes), so far. A long term stability is to be confirmed.
• First, I was planning to use dcdamp at F1 and only dc for the other GAS filters though, is seemed that fro suppressing of 0.2Hz mode, F2GAS looked more effective rather than the others. Then I started to prefer using F1F2-dcdamp, F3-dc configuration.
• During a test, it was found that the BFGAS loop was not closed successfully somehow, due to a high Q peak at high frequency region like 238Hz. This is to be investigated more deeply.

---

When dcdamp is implemented to F1. F2 and F3 GAS:

• For BFGAS, the loop was not closed successfully. It seemed because of a steep line at high frequency like 238Hz(?)
  • [ ] To be investigated more.
• For 0.24Hz and 0.6Hz peaks,those frequency correspond to the dip-frequency of the mechcanical response of F1GAS. They would be removed by feedforwarding to theF1GAS actuator if necessary.
  • [ ] check the necessity

[Note] the spectra are same, just see the right hand side:

---

When dcdamp is implemented to F1, and dc to F2 and F3 GAS:

---

When dcdamp is implemented to F1 and F2, and dc to F3 GAS:

---

Here, I selected "F1F2-dcdamp, F3-dc configuration".

The gain peaking(?) of F2GAS at around 2Hz, is to be taken case of.

[YoshinoriFujii]

For 2019.9.25th 13:18JST condition

• I returned to the initial condition:

[YoshinoriFujii]

data: /users/VIS/TypeA/general/LogNotes/20190924/ITMX

[YoshinoriFujii]

2019.9.26 long-term performance check

At 6:51JST, I set the following filter banks for the stability test:

• At around 7h30, TF measurement at the BF-sateg was done: /users/VISsvn/TypeAtower/ITMX/Measurements/20190926/

• According to TF, the shape was smae as the previous measurement which was doene on Junuary 2019. However, dc loop looked NOT working as expected and high frequency noise (> 500Hz?) looked injected. => why?

• For the time being, the BF-stage loop is kept disabled.

[YoshinoriFujii]

2019.9.26, I finished with