TS: What do you mean by misaligning the S-band structures by 500 μm?
YK: All cells are moved by 500 um in the same direction
TG: But then it's just an injection offset!
YK: At the retreat, I will show a simulation with randomized offsets.
FLP: What diameter do you assume for the tube when calculating the wake fields?
YK: The wakefields are calculated cell to cell...
MP: The misalignement is expected to be of the order of 10-20 μm, should be no problem with respect to wake fields.
MP: How can C-band deal with spikes in the current distribution?
YK: For a smooth profile, C-band is much better than S-band, but if there are spikes in the current profile, C-band does not work so well and S-band is better.
JYR: What if you take only 2 m long S-band structures?
YK: I have not simulated this, but I expect it to be better with respect to sensitivity to misalignment. The transverse wakefield effect, however, will not be very strong at high energy; you need a high wake field to compensate for the energy chirp.
FLP: Why is the C-band structure 2 m long and not, say, 2.5 m?
YK: In that case The klystron could not support the power to drive it! There is only a 50 MW klystron available from Toshiba
FLP: You say the 7.5 GeV upgrade would be easier with C-band, this means that reaching 6 GeV is even easier?
YK: With C-band you have more space in the tunnel, therefore it is easier to reach any energy.
MP: What effect would the C-band solution have on the UV FELs (FEL3)? The chirp compensation should be better than with S-band?
YK: C-band option will be better for UV FEL, but there is not big difference as the chirping compensation is not dominant. For C-band you only need 2 FODO cells, for S-band you need 3 FODO cells.
MP: From the point of view of operation, if you want to compensate the chirp at 2.1 GeV, this means you have to run out of phase.
YK: At 2.5 GeV there is no good way to compensate chirp. Our project is aimed at 1 Angstrom, we should focus on that!
MP: Well I am just wondering if C-band gives us more margin...
FLP: What is difference in price between C-band and S-band?
MP: There is no dramatic difference.
FLP: Could it be that four years down the road, with further maturing of the C-band technology, the difference may vanish?
MP: There is some uncertainty from the low-level RF system.
FLP: Have you thought about cooling the cavities to gain in resistivity?
YK: The operation would be more costly.
FLP: You don't need full cryo, liquid nitrogen would do, and that is cheap.
MP: Still, you will have several kW per m, if you integrate over the full length, this becomes substantial.
BB: But on the other hand you would need fewer klystrons!
TG: Keep in mind that the RF pulse would immediately increase temperature, you would need some flow to carry away the heat...
More discussion on merits and problems of cooled cavities...