As explained in the previous meeting, the diameter of the Y' torsion bar has been changed from 20mm to 15mm in order to reduce the vertical movement of rotational center relative to the geometrical center. Resultant movement was less than 3mm as expected.
Driver circuit of piezoelectric device (PZT) was adjusted, which is a 4ch HV driver with noise less than 2mV rms in the dynamic range from 0 to 150V. The vertical position of the y(Y) mover was measured by a laser interferometer monitoring the fringe conditions by a position sensitive photo-diode. Output voltage of the photo-diode has been calibrated to be 6mV/nm. When the HV driver was ON and OFF, the rms noise was 14.7mV and 50mV, respectively. A the ON, 150mV peak-to-peak noise was also observed at 180Hz. For denoising, a RC filter was inserted with C of the piezoelectric devices which is 7.2uF (30uF) for short(long) device, where R was set to be 2.2kΩ for f=10Hz. So the noise got back to 14.7mV.
The piezoelectric mover system has been checked. Movements in X,Y,X' and Y' were measured by the Quadrant photo-diodes(QPDs). The four movements can be controlled by 4 PZTs, where a reference voltage was supplied by DAC. Also, they can be stabilized by feedback signals from the QPDs. First, the control performances were examined by step function response in the 4 directions. Resultant transition time was measured to be about 100msec in all the directions. Second, back-and-forth scans in the Y direction were measured for almost full dynamic range, while the other positions, i.e. Y, X' and Y', were observed to be stable. Third, the 4 positions have been randomly set and the QPD positions were measured in every 3 second for 3 hours. For an example, the Y positions were measured to be linear to the Y set values, while the other positions were randomly distributed as a function of the Y set values. It shows no cross-talks among the 4 directions. Finally, differences between the set values and the measured ones were plotted for all the directions. They were about 1 bit of ADC (3um) which is also the same bit of DAC.
A new BPM electronics was checked at a test bench. The BPM electronics consists of combiner, converter and detector, which is different from the SLAC one. Common modes in two port signals of the BPM are subtracted at the combiner. The c-band (6.5GHz) signals are down-mixed to IF=700MHz at the converter. Then, the converted signals are detected in phase with the reference signals at the detector in order to output DC signals proportional to the BPM positions. Input signal is a sine wave at c-band (6.5GHz). Dynamic range of the converter (attenuator + amplifier + down-mixer) was measured by changing the input power from 6dBM to -30dBM with variable attenuation for 0-60dB. The output power was "saturated" at 9dBm and -18dBm. The phase detection was tested with the input power of 0, -10, -16 and null dBm, where the reference power is -10dBm, by changing a phase adjusting dial. Also, the output offset dependence on the reference power was measured by terminating the input(Xin) port as a function of the reference input power. The linearity was measured in the intensity output, too.
Laser frame for optical anchor was explained. An elementary unit is a "laser BPM" which consist of prisms and two laser beams in parallel for correction of tilt angle. Thus, the laser BPM can measure vertical movement at one point even if the base is tilted. The laser frame is made by aligning the three laser BPMs on floor and the reference bar. Relative displacement between the floor and the reference bar can be monitored by laser interferometer. Prototype system with three laser BPMs will be constructed in a test vinyl house (assembly hall), where the girder and the reference bar are assembled for off-line test without beam. The prototype system can measure vertical movements on floor separating in distance of a few m.