Reports 1-1 of 1 Clear search Modify search
MIF (ITF Control)
satoru.takano - 1:05 Friday 01 November 2019 (11269) Print this report
FPMI is back!

[Ushiba, Aso, Akutsu, Takano]

We finally recovered FPMI for the first time in 2 weeks.

What we did today
 
Check of stability of CARM loop
We measured OLTF of CARM loop of
  1. ALS only
  2. ALS and IR

then reconstructed IR only by calculation. From this measurement we confirmed that IR CARM loop is stable by itself and 40dB attenuator was too much. Now the attenuator is left removed.

http://klog.icrr.u-tokyo.ac.jp/osl/uploads/11269_20191031170529_carmolg.png

Fig.1 measured and reconstructed OLTF of CARM loop

 
Adjustment of DARM loop gain
After DARM hand off it seemed that DARM is still noisy to lock cavities stably, so we made the gain of DARM loop as big as possible. Finally we increase the gain by 4 times.
 
CARM hand off trial
After that adjustment, we succeeded in hand off CARM signal stably. It seemed that the instability of locking came from lack of DARM control gain.
 
Transition of MICH signal from 51MHz to 17MHz
Next step is transition of MICH signal from 51MHz to 17MHz of REFL PDA1. We did swept sine measurement and decided the relative gain btw 51MHz and 17MHz. Using thea value, FPMI was fully locked.
 
After the first success, we modified the filter condition and reconstructed OLTF as 2 weeks ago. Now OLTFs of DARM and MICH loop are the same as before. For CARM it is impossible to measure overall OLTF because the control loop of CARM contains a fast path (analog circuit).
http://klog.icrr.u-tokyo.ac.jp/osl/uploads/11269_20191031170255_micholg.pnghttp://klog.icrr.u-tokyo.ac.jp/osl/uploads/11269_20191031170246_darmolg.png
Fig.2&3 OLTF of MICH&DARM. blue: 2 weeks ago, red: today
Images attached to this report
Comments to this report:
takahiro.yamamoto - 3:03 Friday 01 November 2019 (11270) Print this report
Noise sources limiting sensitivity also don't change before two weeks ago.

[MICH]
around 100Hz => Intensity noise (MC Trans)
above 1kHz => Frequency noise (MCL, CARM_SERVO_FAST)

[DARM]
above 100Hz => Intensity noise (MC Trans)

Images attached to this comment
takaaki.yokozawa - 3:19 Friday 01 November 2019 (11273) Print this report

Thank you for your great efforts!

I leaved the current sensitivity curve.

http://klog.icrr.u-tokyo.ac.jp/osl/uploads/11273_20191031191913_screenshotfrom20191101031213.png

Images attached to this comment
chihiro.kozakai - 9:55 Friday 01 November 2019 (11276) Print this report

Just FYI, the longuest lock segment is 1256569018-1256574238, 87min.

I attach guradian channel time series.

Images attached to this comment
chihiro.kozakai - 15:41 Friday 01 November 2019 (11285) Print this report

I investigated intensity noise and the time dependence.

From the attached strain whitened spectrogram, the noise floor in high frequency is not stable in time scale of less than minutes.

On the other hand, attached whitened spectrogram of MC trans looks stable for 1 hour. Therefore, fluctuation comes from other source than intensity noise.

I also attached coherencegram between MC trans and strain. Comparing with the strain spectrogram, the coherence seems small when starin is noisy. Therefore, I suspect that time-dependent more important noise source exist somewhere else. To be investigated.

Images attached to this comment
takaaki.yokozawa - 15:57 Friday 01 November 2019 (11287) Print this report

There would be contamination from the MICH control signal.

Search Help
×

Warning

×