It is possible to obtain a better decoupling of DARM/CARM vs MICH/PRCL/SRCL by using a double demodulation technique. If one of the sideband pairs is created by amplitude modulation instead of phase modulation, it is possible to demodulated at the difference between two frequencies to extract control signals which are independent of the carrier field. Therefore these signals should contain a much lower contribution from CARM and DARM, since the sidebands are not resonant in the long cavities. It is however no more possible to extract useful control signals from the beating of the amplitude modulated sidebands with the carrier.

Assuming that SB1 are still phase modulation sidebands (they are needed to control CARM and DARM), there are three possibility:

• SB2 and SB3 are both amplitude modulated. Therefore one can extract signals from the beatings SB1-SB2 and SB1-SB3, but not from SB2-SB3 or from CAR-SB2 and CAR-SB3
• SB2 is amplitude modulated and SB3 is phase modulated. Signals can be extracted from SB1-SB2, SB2-SB3, CAR-SB1, CAR-SB3 but not from SB1-SB3
• SB2 is phase modulated and SB3 is amplitude modulated. Signals can come from SB1-SB3, SB2-SB3, CAR-SB1, CAR-SB2 but not from SB1-SB2

The last option seems less interesting to obtain good error signals to control SRCL.

The second option only is analyzed here, since it allows to maintain signals coming from the beating of carrier and SB3, which might be useful for CARM.

• Short Schnupp asymmetry - 11m long signal recycling cavity

• Short Schnupp asymmetry - 12m long signal recycling cavity

• Large Schnupp asymmetry