Akutsu and T. Yamamoto (for safety) : continued from 10321.

Measured the profiles of the transmitted IR beam at TMSY along with changing the distance between RLNS1 and RLNS2 several times. For all the measurements, an IR mirror (I found a PYD-10, not the BS mentioned in 10321) was put to fold the optical path to obtain data points after z=500mm. The tendency did not change so much...

• I moved RLNS1 slightly to RLNS2, then got Fig 1.
• Then I also moved RLNS1 slightly to RLNS2, then got Fig 2. It seemed the two fitting (1: using all the data point, 2: using only the first a few data points) were kind of matching for horizontal and vertical respectively, so I expected the strange evolution could be resolved by reducing the distance of RLNS1 and RLNS2 more.
• So I again moved RLNS1 slightly to RLNS2, then got Fig 3. My expectation was not fulfilled...
• I slightly move RLNS1 back, i.e. increase the distance btwn RLNS1 and RLNS2 so that the situation could be closed to Fig 2, and obtained Fig. 4; this is the current status. Fig 5~9 were obtained at z=50mm, 300mm, 400mm, 500mm, 650mm (around the waists), and 900mm respectively. As mentioned already, the beam shape flipped in-between z=400 and 500 (Fig 7 and 8).

As I mentioned, the IR mirror was located to obtain the data points after z=500mm, but that does not mean the IR mirror was always fixed. Somewhat ineffective, but the actual procedure was as follows; I measured the data points from the lower z value along a single optical path to z=450 (=18*25), then put the IR mirror at z=425 (=17*25) to fold the beam path, and then measured from z=500 (=20*25).

Occasionally with Nagano-kun's comment (10329), I slightly suspected the IR mirror, and folded the optical path at difference point to see the beam profiles, and I found more complicated facts.... the beam shapes seemed varied. I repeated the measurements several times to confirm it... hmm... maybe I would like to put a PBS or something in the optical path to confirm that.

... Anyway, this is just a DC GPT, so probably I don't need to mind so much, and I would like to do just putting the QPDs.

Note that the TMSY is still in air, and the gate valve with a window is shut.

... Do we need a 2D Gaussian fit?? or hoepfully fitting also with the wavefront phase... so that we can extract pure TEM00 components.