Design of the reference frame was updated with detailed designs of active feedback mover system for three cavity-BPMs provided by Y. Honda. Overall dimensions are 240mm(width) x 530mm(height) x 1160mm (length). L-shaped head is made of 88kg Aluminum, which is supported by 4 legs of SUS(37kg). Aluminum was chosen for lightweight. However, the high thermal expansion coefficient (2.35x10-5/deg.C) was pointed out at this meeting. Typical temperature variation is +/- 1 deg.C at the ATF. So, Invar (super invar) would be best for 9(1) x10-7/deg.C . ( Granite has 7 x 10-6/deg.C .) An effect of air flow was also pointed out. The reference frame and the mover system are bolted on a SUS plate of 40mm thickness (226kg), which is put on a granite table of 1000x1500x200mm3, 900kg. The granite table is supported by iron frame made of channel(100mm width x 50mm height). The total height can be adjusted by level jacks anchored at floor. It was pointed out that the iron frame could be replaced by three granite beams for stability.
The total system was analyzed by FEM model. Deformation due to self-weight was estimated to be 0.02mm on top of the reference frame. Resonance frequencies were also calculated to be 40, 53, 66 and 87Hz at 1st, 2nd, 3rd and 4th resonances assuming all the connections are rigid like as welding. Bolted connection may decrease the frequencies. Spectrum analysis was also applied with ground motion measured at ATF, Sep.1st,noon. The system follows the ground motion below 10Hz, while the power spectrum density was peaked to 10-19(17)m2/Hz in vertical (horizontal) direction at the 1st resonance.
An optical fiber of kineFLEX was tested for stability of intensity, polarization and coherency with temperature variations. This fiber showed good stability while output light profile is slightly elliptic. So, two fibers were ordered.
A prototype of active feedback system was fabricated. The prototype has a circular base plate supported around by elastics hinge (0.1mm thick and 5mm width), while the base is mounted on a piezo device(PZT). A dummy load of 3kg, which is same as the cavity BPM's weight, can be put on the base. First, a mechanical resonances were measured to be 4kHz and 1 kHz without and with the load, respectively, by a capacitive sensor, which are high enough for any disturbance motion. Then, an active feedback was tested by using a laser interferometer, where the laser was split into two toward a mirror on the same optical plate as the laser and a second mirror glued on the dummy load. A fringe pattern of the interference was measured by a photo diode segmented into two ( A and B). Error signals of A-B were integrated into step ones for amplifications with gain adjustments. The amplified signals were applied to the PZT for active feedback with the step response of less than 100Hz. A lock-in stabilization was demonstrated against 35 ~ 60nm fluctuations with the RMS of 0.35nm for about 100sec .
A stacked piezo housing prototype was also tested with and without a 50kg load. Positions were changed for 100 um with applied voltages from 0 to 90V. The position change was about 10um with the load in the voltage range, Clear hysteresis was observed in different voltage change, while a good stability was also observed within 1 um for a few hours.
A concern on the brittleness was pointed out. So, a moisture control must be important. Also an endurance test might be necessary since it would be difficult to replace them once they are built in.