1. Abstract The Hannover Thermal Noise Experiment V. Leonhardt, L. Ribichini, H. Lück and K. Danzmann Max-Planck- Institut für Gravitationsphysik We measure the differential movement of two mirrors suspended as multiple stage pendulums to detect the thermal noise of the pendulum mode. The setup and different efforts to reduce some noise sources are presented here. Institut für Atom- und Molekülphysik Optical Setup The frequency of the laser is stabilized to a reference resonator with the Pound- Drever-Hall scheme and the length of the suspended resonator is stabilized to the laser frequency. Below the unity gain frequency of a few hundred Hertz, the feedback signal of this loop is proportional to the differential movement of the suspended mirrors, above unity gain the error signal is monitored. To achieve a constant alignment of the laser light with the suspended resonator, an auto alignment systems centers the laser beam on the resonator mode. Blades Double Pendulum Damping Double Pendulums Piezo Mirror Vacuum Chamber Active Seismic Isolation Frame Mirror Laser Commercial Active Isolation The mirrors are currently suspended with four pendulum stages and two blade stages. The seismic excitation is reduced further by an active seismic isolation system. The pendulum resonances are damped at two stages with a coil-shadow sensor-magnet system. The 3cm long optical resonator is locked to the laser frequency by a piezo, but it can also be locked by a coil-magnet system. Seismic Isolation Sensitivity The sensitivity in measuring the differential pendulum movement is limited below 50 Hz by a combination of seismic noise and feedback noise of the control loops damping the pendulum resonances. The seismic noise around 10 Hz is not limited by longitudinal movement, but by coupling of the other degrees of freedom into the longitudinal. To minimise the influence of the other degrees of freedom we have increased the moments of inertia of the pendulum masses by adding small wings with heavy masses at the end. This reduced some resonance frequencies and enhanced our sensitivity. However another attempt with even bigger arms attached to the pendulum stages was unsuccessful as the number and strength of internal resonances increased too much. Conclusion: Several noise sources have been identified and addressed, however the achieved sensitivity does not yet allow the measurement of the pendulum thermal noise. Noise 1000 10000 Frequency [Hz] 10 -17 10 -15 Movement [m/Hz 1/2 ] errorpoint electronic noise errorpoint sensitivity Measurements with a fixed cavity, eliminating the influence of seismic and other noises moving the pendulums have shown that the seismic influence is only dominating the sensitivity at frequencies below 50 Hz. Above that a combination of different electronic and laser noises prevent a better sensitivity. Using mirrors with a transmission of only 100 ppm a better signal to noise ratio at the photo diode and the mixer was achieved which improved the sensitivity in the kHz region. 10100 Frequency [Hz] 10 -19 10 -17 10 -15 10 -13 10 -11 Movement [m/Hz 1/2 ] current driver diode+mixer rpn (callibrated with aom) ampl. noise in pdh signal frequency noise (in loop) fixed cavity pendulum thermal noise sensitivity