This note relates to how the existing steering package in the SCP maybe used to implement eta correction using only an antisymetric corrector terms, and therefore avoid effects of dispersion correction spilling into orbit (question).

When steering dispersion, the system matrix that is minimized by steering specifically does include the 2nd order quad and sext terms (some people thought it didn't).

So, if you initialize from rmats, and you do x&y orbit steering, then since the orbit minimization is typically much more highly weighted than the dispersion correction, the cross plane coupling is taken into account at least for orbit. It's true that, to the extent that a correction causes an off-axis orbit through a skew-quad, the dispersion induced in the opposite plane is not taken into account.
For this reason, Uli wants (possibly "wants" will become "wanted" - see below) a project to create a method of doing orbit and dispersion correction in which the variables of the minimization are not the individual correctors but "knobs", or linear combinations of correctors that they have devised to change the orbit but always steer through the center of skew-quads. That is, they're asymmetric "4-bumps" centered on a skew-quad. Steering would not then use any individual correctors but only these knobs.

So there are few ways one could do that. We may be able to use the existing software using the "offset" and "weight" options on the steering's BPM setup panel to explicitly add constraints for the orbit through skew-quads. This means someone would have to write something that acquires an orbit (aida even!), and then write a button macro to set the offset and weight for the right BPMS. Uli thought that was worth an experiment.

  • No labels