KAGRA Logbook

AOS (Beam Reducing Telescopes)

koji.nagano - 21:44 Thursday 12 April 2018 (4692)

Fluctuation of green light at TMSY reflects ones of PR3 and BS

[K. Nagano and Akutsu]

We measured the coherence between the output signals from QPD1 at TMSY and OpLevs of PR3 and BS as shown in left panels of the following picture.
Please note that TMSY QPDs have not been calibrated!
As you can see,

PIT: coherent to PR3 between 0.5 and 20 Hz (dip at 3.8 Hz comes from notch filter in PR3 IM OpLev control),
YAW: coherent to PR3 at 1.3 Hz and to BS at 0.47 Hz.

In addition, we measured the spectrums of QPD1 signals and PR3 OpLev PIT.
Please note again that TMSY QPDs have not been calibrated!!
As you can see,

PIT: the shape is similar to PR3,
YAW: the peaks at 1.3 Hz and 0.47 Hz from PR3 and BS, respectively, are exist. Moreover, there is the dip at 3.8 Hz (maybe PR3 PIT control is mixed to yaw motion somehow).

This figure tells us that, to reduce the fluctuation of the green light (or any light reflected by PR3 and BS, e.g. main IR beam), we should damp PR3 and BS.

By the way, why can the dip at 3.8 Hz be seen in the TMSY QPD1?
Is the control loop UGF of PR3 PIT with OpLev around 10 Hz?

Images attached to this report

Comments to this report:

kiwamu.izumi - 4:11 Friday 13 April 2018 (4698)

Hi Nagano-kun,

Thank you for looking into this. Here are my answers to your questions.

The bandwidth of the PR3 optical lever loops are supposed to be 3 Hz for both pitch and yaw (3828 and recently Yutaro modified the DC part of it only for the pitch loop; 4648). You can sort of guess how high the bandwidth might be by staring at the differences in the spectral shapes of the PR3 optical lever and the TMSY QPD readout.

It seems that the coherence at 3.8 Hz is low because that is the frequency in which sensing noise of the PR3 optical lever is NOT fed back to the optic (4648). This strongly indicates that the PR3 motion is limited by some broadband sensing noise imposed to the PR3 optic via the feedback loop above 0.4-ish Hz. This is something Yutaro was worried when he was modifiying the PR3 control loops. If you want to reduce the noise contribution from PR3, you can consult with Yutaro for reducing the PR3 control bandwidth. Or you might want to consider reducing the sensing noise of the PR3 optical lever somehow.

By the way, I have a question for you just out of my curiosity: To what physical quantities do you mean to calibrate the TMS QPDs?

koji.nagano - 9:18 Friday 13 April 2018 (4700)

Hi Izumi-san,

I uderstand the situation of PR3 control.
I talked with Yutaro and we will try to reduce the PR3 control bandwidth at some point.

> "To what physical quantities do you mean to calibrate the TMS QPDs?"

I did not consider this so seriously.
Calibration to the beam position on TMSY QPD surface may be OK at this point, i.e. while TMS is not used for global angular sensing and control.
The beam poisition may be not so meaningful though.
What I wanted to say in my original post, the absolute value of TMSY QPDs did not have physical meaning.

koji.nagano - 14:52 Friday 13 April 2018 (4704)

I checked PR3 OpLev's wind shield status.

Transmitter: Have a wind shield.
Receiver: NOT have wind shield.

Installation of a wind shield to PR3 OpLev receiver will help to reduce the sensing noise.

I attached a not so good picture.

Images attached to this comment