Yuta, Kiwamu, Koji, Stefan - Yuta's calculation (see other alog) showed that the error signal significantly depends on the modulation frequency. - We therefore used and IFR to try several choices of frequencies. - We measured the FSR at 24.429380MHz +-20Hz, and 24.391480MHz +-20Hz - This gives us an arm length of 3992.973m +-6m.... - The estimated transverse mode spacing is 5.12kHz. (R_ITM=1939.52m, R_ETM=2241.54m) - We tried to set the frequency to exactly FSR +- 5.076kHz (an old estimate of the transverse mode spacing), as well as half of that. Subjectively the best performance (least mode hopping) was at 24.424304MHz. It was not perfect. Still to do: - Find a spot with minimal locking point dependence on the 10-mode power. - Explore the I/Q demod phase.