Kuroda made lists of position (dx,dy) and field-strength (dk/k) errors for quadrupole magnets from QD0 to QD10A in the final focus system on request from Yokoya-san. As in the previous meeting, each error generates 1.5 times beam sizes and displacements of x*,y*=&sigmao*x,y , where &sigmao*x, y are 200 and 4 nm, respectively. As expected, the final doublet (QF1 and QD0) has the smallest position error at nm level in vertical direction. The displacements always require smaller errors at all the quadrupoles except for QF7 and QD2B. The smallest error of dk/k was estimated at the final doublet too, which was 2.41 (9.29) x 10-5, while the errors are more than 10-4 at other magnets. These errors shall not be static since the static ones are 0.2~0.3%. Actual tolerances of the stability will be estimated together with feedback and correction schemes.
We also discussed on the permanent magnet option for the final doublet especially concerning of field variations from temperature change and radiation damage etc. .
(4) FEATHER (N.Delerue)
(transparencies, 6 pages, pdf, 365KB, and a new beam test result PDF, 2 pages, 87KB)
Since Nicolas had a midnight shift for the beam test, Tauchi introduced his transparencies. First, the transparencies show results of BPM calibration in previous beam test and calculation of BPM signals as a function of gap. At the previous beam test he could not calibrate the BPM, while he finally succeeded it in this midnight shift. So, the result should be replace with the new result (pdf) which was sent in this morning from him. The calculations were found to be wrong because of a bug in the program, and they are not consistent with this new beam test result. These updates will be reported at next meeting.
Nicolas also estimated effect from the beam blow up due to small crossing angle in the fast feedback system. He concluded that "if we kept memory of past beam position in delay loops (one delay loop per past bunch), we should thus be able to add this additional effect in our corrections without problems. Efficiency of the system is the same than for the improved model (very similar)." However, the simulation results on the feedback performance should be expected with the beam blow up.