Linearity of the Electron Optics - PowerPoint PPT Presentation

1 / 16
About This Presentation
Title:

Linearity of the Electron Optics

Description:

In many cases, YAG images of e-beam had distinct triangular shape, ... For a linear local 3-bump around a doublet. A nonlinear offset in the bpm BS4 , for ... – PowerPoint PPT presentation

Number of Views:67
Avg rating:3.0/5.0
Slides: 17
Provided by: beamsdi
Category:

less

Transcript and Presenter's Notes

Title: Linearity of the Electron Optics


1
Linearity of the Electron Optics
Alexey Burov
RR Talks
2
Examples of YAG images
  • In many cases, YAG images of e-beam had distinct
    triangular shape,
  • pointing to a sextupole nonlinearity (Sasha,
    Lionel).
  • This nonlinearity has to be seen in differential
    orbit measurements.
  • The differential orbit measurements were done by
    Mary
  • A5 A6 correctors Oct 29 (not clean in CS) ,
    Nov 6 (OK for CS)
  • S2 S5 - Nov 1
  • S3 Nov 26 S4 Nov 29.

3
Nonlinearity Coefficient
  • For every mentioned corrector, 10 equidistant
    kicks were applied, from the highest possible
    negative, to the highest positive. For every
    corrector j, the BPM data were saved in P163
    format.
  • To see nonlinearity in these data, I prepared a
    MathCad file, which makes a linear fit for every
    given BPM i , and subtract this fit from the
    data
  • For every BPM i, the nonlinearity is
    characterized by
  • For every corrector, the measure of nonlinearity
    of its trajectories can be presented as

4
Example for CXA05
BXS05, cm
BYS05, cm
5
E-Cool Line
6
Results
  • The numbers are not negligible. Currently, the
    angle is estimated 100µrad on the axis.
  • Nonlinearities for below CS2 are small.
  • Also nonlinearities for BS2, BS3 were always
    small (not shown here).
  • Max nonlinearity are always at XC4, YC8 ( p/2
    of the Larmor phase).
  • Hypothesis
  • The main source of nonlinearity is located at
    the lens S3

7
Accuracy (CYS2 kicks)
Error of CYS2 current (or is it MI perturbation?)
is equivalent to 100-200 microns of the bpm
error. And this case is not unique How can we
avoid this?
8
How the error was evolved downstream the line
9
How to check the hypothesis
  • To check the SS3 hypothesis, a local bump
    CYS2-CYS3-CYS4 can be done.
  • Then, for several bump mults, bpm data can be
    taken and compared with the nonlinear parts of
    Nov 1 CYS2 measurements.
  • Nov 1 CYS2 CXS2 measurements should be redone,
    due to their high errors for some points.

10
Cylindrical Aberrations of Ideal Lens
  • Focusing strength of a solenoid can be derived
    as
  • Assuming
  • In local bumps, this nonlinearity has to be seen
    as a cubic parabola in a plane of the bump, and
    nothing in the other plane, as soon as the lens
    is optically thin.
  • For

11
Bump nonlinearity
  • For a linear local 3-bump around a doublet
  • A nonlinear offset in the bpm BS4
    , for

SS3

-
BS4
BS3
CS2
CS3
CS4
12
Cx/yS2 Bump Data
CYS2 bump
CXS2 bump
Huge nonlinearity of SS3! 14/2.56 times of the
ideal level
13
CS2 bumps show
  • Nonlinearities seen in these bumps
  • None of them has symmetry of the solenoidal lens
  • Y-bump is 3 times more nonlinear as X-bump (as
    in P163, p.5)
  • Absolute value of Y-bump nonlinearity is 5 times
    higher the ideal level (also agrees with the P163
    data in p.5).
  • Dependences of BS3(CS2) are all linear at the
    noise level.
  • Thus, these bump data confirm my hypothesis about
    extremely nonlinear S3 lens

14
CS3 bump
CYS3 bump
CXS3 bump
Nonlinearity of SS4 is 5/2.5 2 times of the
ideal level.
15
Lens T4
  • The same way of measurements with the 3-bump
    around ST4 (Mary, 08Feb21_linearity measurement)
    yields non-linearity as 4 times of the ideal
    level.

16
Conclusions
  • Bump measurements give the same ratios of
    nonlinearities as P163.
  • The hypothesis of extreme SS3 nonlinearity (6
    times more than natural , 3 times more than SS4)
    is confirmed.
  • Other measured lenses have following
    nonlinearities
  • SS4 2 times of the natural level
  • ST4 4 times of the natural level
  • It is reasonable to explore the most linear part
    for the beam position in SS3 (-5 mm), and to
    optimize optics, reducing the field in SS3.
Write a Comment
User Comments (0)
About PowerShow.com