KAGRA Logbook

MIF (General)

kentaro.komori - 23:41 Wednesday 01 February 2017 (2425)

Derivation of the sensing matrix on WFS

[Kokeyama, Komori]

We are keeping working on WFS in order to derive the sensing matrix on WFS.
When I-phase demodulated signal of each QPD (A1 and A2) is maximum, whose phase was reported yesterday,
the gain [count/rad] is as below.

YAW

MCI MCO MCE

A1 -11000 -1440 -5060

A2 -7410 -1010 -5070

PITCH

MCI MCO MCE

A1 684 -299 -4920

A2 -585 105 6530

Results of yesterday were corrected and MCE was added.
The gain of common mode of MCI and MCO (shift) and differencial mode (tilt) as for YAW are culculated as follows.

shift tilt

A1 -9530 -1240

A2 -6400 -8420

We did same measurement when the demodulated phase is all 0 degree.

YAW

MCI MCO MCE

A1 -8760 -1220 -4430

A2 -2760 -377 -3840

PITCH

MCI MCO MCE

A1 585 -200 -5360

A2 -535 32.7 -1202

shift tilt

A1 -7530 -9980

A2 -2390 -3140

Now we are considering how to get signal of shift and tilt mode from A1 and A2 signal.
The problems are the reason why demodulated phase is totally different depending on each quadrant,
non-centering of beam spot on QPDs and that WFS signal of yaw and pitch mode is coupled.

Comments to this report:

keiko.kokeyama - 10:05 Friday 03 February 2017 (2431)

It was good we could make some whatever measurement of wfs in the limited time!

This time we could not tune even demodulation phasese for each quadrant, and I think we must re-shape the beam profile of the IMC REFL path to make the better measurements. Because the beam size at each WFS are too small (~400um in radius), and all the quadrant cannot receive the laser light all the time. I think this is why we cannot tune the demod phases. The beam spots at the WFSs are moving too much and it also makes impossible to set the demod pahse. It can be expained also by the too small beam spots (too diversing beam). I believe this point was not taken into account when the beam profile of IMC REFL were designed. Also the actual beam shape is different from the design, in anyways.

We'll wait for the new PSL to come in April to refine the configuration, and to make the WFSs work.

keiko.kokeyama - 10:11 Friday 03 February 2017 (2433)

By the way, I think Komori-kun meant [cts/urad], not [cts/rad].

Also, I wonder if he used just the heights of OPLEVs PIT/YAW spectra to derive the amount of excitations [urad], or used the RMS areas of the spectra?

kentaro.komori - 9:09 Tuesday 07 February 2017 (2447)

I'm sorry for my late response.
The unit is cts/urad, and I used the peak heights of spectra, not RMS areas.