Confusion to the RFMODE element

[michael_borland]

Jiang,

The problem is that your RFMODE-based simulation uses only a single bunch, which will have large transients and compression from the cavity wakes. To solve this, you can use pseudo-multi-bunch mode, which involves tracking one particle distribution but driving the cavity fields as if you had a series of equispaced bunches. The attached files show how to do this. The result is good agreement of the two methods.

As for the phase walk-off, this is a problem with the feedback implementation. The beam and rf systems walk off together, locked in phase, but drifting relative to the absolute clock. We don't have a ready solution now, but it seems harmless.

--Michael

[michael_borland]

Forgot the files.

--Michael