There are four sets of plots at the bottom of this page (each of them with two plots). The data used to generate them are the current feed-forward tables for the parallel or circular-polarization mode (as used on 2002-10-01 at the ALS at 1.9 GeV). The upper plot in each of the images shows the horizontal plane, the lower plot the vertical plane. All kicks are in rad, please be aware that the axis scaling is different in every plot in there is a scaling factor (1e-5 or 1e-6) on the top left corner of every plot.
The first two images are for the EPU in sector 4 (at a minimum gap of 14.9 mm), the last two are for the EPU in sector 11 (at a minimum gap of 15.67 mm). The first and third image shows the kick of all four corrector magnets used in the feed-forward (hor./vert., upstream/downstream). The chicanes are significantly closer to the EPU than the corrector magnets.
The second and fourth image shows in each of the planes just the sum of the two kicks (to make it easire to compare it with Andy's simulations). One should remark that depending on the exact longitudinal distribution of the kick from the EPU the sum of the two corrector kicks might be somewhat different from the direct field integral (due to the very nature of the closed orbit). But it should be a fairly precise estimate.
Conclusion: Measured kicks in the vertical plane for the two devices in sectors 4+11 are in reasonably good agreement, but are a factor of more than 5 larger than the ones initially calculated in RADIA (calculation for ideal device including the ends and the deviation of dB/dH from 1). Oleg Choubar later found that the slicing in our Radia model was not fine enough. With suffieicent slices the simulations agree within 10% with the measurements (Oleg's simulations will be posted soon).
