To find a good Si position, we need to follow a procedure similar to this:

1. Si in max-in position
2. find distance of beam by scraping with detector from inside 
--> zero position
3. from this zero, we need to set a distance of about 1.5 cm to the inside (as shown in the sketch below)

It is likely that this position is not compatible with the complete ESR cycle (e.g. we scrape beam with det. during deceleration).
The solution is to move the beam closer to the detector after deceleration. There are two feasible methods to do this:

A. make a local bump in the dipole
--> check new zero and det. position again, as sketched above
--> in 2016 this method didn't allow to go close enough to the detector, eventually we used:

B. global orbit change (by magnet ramp) + target bump (to keep overlap)
--> this takes a bit more time to set up and also the cycle will be longer at low energies (critical due to life-time)
--> method A is preferred, but might not be strong enough

This needs to be checked and repeated for each new beam setting.