Significant improvement has been achieved in the RoC measurement.
By these changes, dramatic increase of the signal to noise ratio was seen.
Now both of the peaks corresponds to the 1st-order higher-order modes are clearly seen.
The peak at around 26MHz are produced by the beat between the carrier TEM00 and the upper-sideband TEM01 (or 10).
The other peak at around 57MHz are produced by the lower-sideband TEM01 (or 10).
From the peak fitting we can extract the following numbers:
Note that the cavity itself has not been touched during the measurement.
Only the laser frequency and the incident beam alignment were adjusted.
The results are calculated by the combination of MATLAB and Mathemaica. The fit results are listed in the PDF files.
In deed the fitting quality was not satisfactory if the single Lorentzian peak was assumed.
There for two peaks closely lining up with different height. This explained slight asymmetry of the side tails
This suggests that there is slight astigmatism on the mirrors (why not.)
The key points of the results:
- FSR and the cavity length: 83.28~83.31MHz / L=1.799~1.800 [m] (surprisingly good orecision of my optics placement!)
- Cavity g-factor: Considering the flatness of the flat mirror from the phase map, the measured g-factors were converted to the curvature of the curved mirror.
RoC = 2.583~4 [m] and 2.564~7 [m]. (Note: This fluctuation can not be explained by the statistical error.)
The mode split is an order of 10kHz. This number also agrees with the measurement taken yesterday.
If the curved mirror had the nominal curvature of 2.5m, the flat mirror should have the curvature of ~20m. This is very unlikely.
- Approximate cavity line width: FWHM = 70~80kHz. This corresponds to the finesse of ~500. The design value is ~780.
This means that the locking offset is not enough to explain the RoC discrepancy between the design and the measurement.