The first attempt not to touch the curved mirrors did not work. (Not surprising)
The eigenmode is not found on the mirror surface.
We decided to touch the micrometers and immediately found the resonance.
Then the cavity alignment was optimized by the input steering mirrors.
We got the cavity length L and f_TMS/f_FSR (say gamma, = gouy phase / (2 pi) ) as
L=1.1347 m (1.132m nominal)
gamma_V = 0.219176 (0.21879 nominal)
gamma_H = 0.219418 (0.21939 nominal)
This was already sufficiently good:
- the 9th modes of the carrier is away from the resonance 10-11 times
of the line width (LW)
- the 13th modes of the lower f2 sideband are 9-10 LW away
- the 19th modes of the upper f2 sideband are 1-3 LW away
This seems to be the most dangerous ones.
- The beam spots on the curved mirrors are too marginal
So we decided to shorten the cavity round-trip 2.7mm (= 0.675mm for each micrometer)
and also use the curved mirrors to move the eigenmode toward the center of the curved mirrors.
After the movement the new cavity length was 1.13209 m.
The spot positions on the curved mirrors are ~1mm too close to the outside of the cavity.
So we shortened the outer micrometers by 8um (0.8 div).
This made the spot positions perfect. We took the photos of the spots with a IR sensor card.
The measured cavity geometry is (no data electrically recorded)
L=1.13207 m (1.132m nominal, FSR 264.8175MHz)
gamma_V = 0.218547 (0.21879 nominal, 57.8750MHz)
gamma_H = 0.219066 (0.21939 nominal, 58.0125MHz)
- the 9th modes of the carrier is 11-13 LW away
- the 13th modes of the lower f2 sideband are 5-8 LW away
- the 19th modes of the upper f2 sideband are 4-8 LW away
The raw transmission is 94.4%. If we subtract the sidebands and
the junk light contribution, the estimated transmission is 97.6%.
Even if a mirror is touched (i.e. misaligned), we can recover the good alignment by pushing the mirror
onto the fixture. The fixture works pretty well!