We had the 7th IP-BPM meeting at 10:00-11:00, 3rd August 2006, 222, 3-gokan, KEK.
Participants: P.Doublet, Y.Honda and T. Tauchi
His presentation is here.
Estimation of x-y isolation in the electronics
The x and y, i.e. 0 and 90o signals, isolation was studied in the detector (electronics), Measured signals are plotted as a function of phase adjustment dial with 240 units/180o. The isolation was estimated to be -40.4dB/unit (-37.9dB/deg) while the analytic one was -39.0dB/unit (-36.5dB/deg). Suggestions and comments are made as follows; (1) add measurement errors on data points, (2) add the theoretical curves for comparison, (3) additional errors may come from phase adjustment by beam.
Cavity design for an ILC-like beam (3)
In order to check his calculation, the IP-BPM was re-drawn by MAFIA with only rectangular shapes of Honda's dimensions. S parameters of transmission(S11) and reflection(S12) were calculated with perfect conductor. Also, Qexts were calculated by f/Δf in both cases. In general, the perfect conductor must have transmission(S12)=1 and reflection(S11)=0 if the wave guides are optimized. However, his results were S12=0.36 and S11=0.64 . The unexpected results may come from different modes excited in the wave guide. The modes can be adjusted by the port position (Honda's comment). During discussion, he found an error in calculations of Qext in the reflection, while Qexts are right in the transmission. Both Qexts must be identical. Calculated Qext agrees well with measured one in y direction, while there is significant difference in x direction. The difference could be due to resonances in the wave guide.
Dimensions of the ILC-like BPM have been determined. The slot widths become significantly larger as shown. He succeeded to obtain Qext=384 in the y transmission, while Qext was too high in the x transmission. Here, there is the same error in the reflection.
He stopped the design for his internship period at KEK. There is no study on the X-Y isolation, and the resultant dipole frequencies became to be different from the design values. He noticed V.Vogel's design with x-band frequency for shorter decay time maintaining "high" Q values for &tau=Q/&omega. Also, he pointed out that the 2nd common mode must be important for it can couple the slot as the dipole modes. It can only be suppressed by a bandpass filter and the combiner .
He is finalizing the report for the internship at KEK.
Finally, he expressed gratitude for many kinds of support at KEK.