I. Vasserman

1
LCLS First Undulator Prototype Magnetic
Measurements

• I. Vasserman

2
Outline

• Magnetic measurements key measurements and
  issues
• Mechanical stability
• Temperature effects
• End phasing
• Summary

3
Critical Tolerances (per undulator section
undulator and quadrupole)/Achieved
Trajectory walk-off ?x?, y 2?m 0.5?m
A/A0 1 2 0.1
?-2?n 10 deg
?y 50?m
Complex Amplitude of Radiation
Phase slippage over the length of one undulator
section
4
Magnetic Measurements (key measurements and
issues)

• Coil measurements
• Moving coil used as reference (especially for
  horizontal field)
• Field integrals (no multipole components)
• Hall probe measurements
• Both vertical and horizontal field
• Fixed gap
• Easier to tune field integrals and phase errors
  (no gap dependence)
• Small gap ( 6.4 mm)
• Shimming more complicated

5
Magnetic Measurements (key measurements and
issues), continued

• High magnetic field stability
• Very precise measurements needed
• Reproducibility of measurements must be ltlt
  required precision of the field ?Beff/Beff 1.5
  x 10-4 ( 2 Gauss 1 µm gap change)
• Earth field 0.15 Gauss --gt 2.0 µm trajectory
  offset (requirement)
• Obtaining the exact field is a challenge
• To obtain the real trajectory environmental
  field in the tunnel has to be taken into account

6
Hall Probe Horizontal Field Measurements

• Test of Hall probe horizontal field measurements
  using Sentron Hall probe done for Undulator A in
  1998 showed good agreement with moving coil
  reference measurements. It means that planar Hall
  probe effect and cross talk between two sensors
  is not so big
• LCLS undulator is longer by 1 m and has stronger
  vertical field that exaggerates the errors of
  measurements. A test was done to check the
  horizontal field readings of a Sentron probe in
  presence of vertical field
• Results show that for small perturbations and
  small horizontal field this probe could be used
  for tuning but final trajectory measurements
  should be compared with reference done by moving
  coil

7
Horizontal Field Nonlinearity vs. Vertical Field

• Measurements done with 2-axis probe in
  calibration magnet
• The angle ? 1 degree was introduced to have X-
  component of
• magnetic field
• BxBy ?
• Hall probe sensitivity 5V/T
• Cross talk is evident
• BxBy ?

8
Table II. Magnetic Measurement Parameters
achieved

• Absolute Hall probe
• calibration accuracy 0.5 Gauss
• Reproducibility
• - Particle beam angle at exit
• and entrance 2.5 G-cm/0.001 mrad
• - Displacement at exit and
• entrance 3400 G-cm/ 0.0004 mm
• - Beff RMS error 0.15 Gauss
• - Phase error 0.02 degree

9
Shimming

• Novel trajectory shims
• Phase shims
• Mechanical shims

10
Shims .phase (flat) and trajectory (side)
Both types of shims
Trajectory shims
11
Temperature Dependence

• Accurate measurements of temperature dependence
  of Beff need to take into account temperature
  dependence of Hall probe
• Temperature dependence of recent Hall probes
  typically lt10-4 / C?
• Two of three Sentron probes (S/N 157 and 367 at
  APS) have temperature dependence close to
  specified
• Result of calibration for third Hall probe (S/N
  409 using the APS calibration magnet) shows large
  deviation from vendor data
• (?Beff/Beff)/?T appears to be 3.0 x 10-5 /?C if
  Hall probe temperature dependence is neglected
  (one calibration file is used)
• (?Beff/Beff)/?T -5.5 x 10-4 /?C when Hall
  temperature dependence is taken into account
• ?T 0.3 ?C will result in ?Beff/Beff
  requirement of 1.5 x 10-4)
• Undulator end-phase corrections will relax the
  temperature stability requirement

12
Hall Probe Temperature Calibration
Coefficient 1.0021 was applied to the curve of
26.85 deg to coincide with 23.2 degree The shape
is close and one coefficient could be used for
each temperature
13
Time for undulator to reach thermal equilibrium

• Temperature response at downstream (D/S) and
  upstream (U/S) end of prototype core and nearby
  air in the magnetic measurement laboratory

It takes one hour to stabilize the room
temperature More than 24 hrs is needed for
titanium core
14
Mechanical Stability

• The prototype was removed from the bench and
  moved twice around APS storage ring.
• Prototype was aligned and measured at the bench
  before and after being moved

Beff (Gauss) rms (Gauss) of Beff
Before move (23.5 ?C) 13724.7 0.1
After move (23.6 ?C) 13724.3 0.22
15
Undulator End Phasing

• Full range is 0.1 mm
• Implemented into the design
• Sensitivity to field deviations from one
  undulator section to the next can be made lower
  by using the end-gap correction system
• Remotely controlled at the sub-micron level

PZT translator located at the end of the
undulator to adjust the magnetic gap of the end
section. This adjusts the phasing between
undulators
16
Undulator End Phasing (contd)

• End-phase corrections effect on the FEL
  performance
• Calculations of complex amplitude of radiation
  amplitude
• Simulations of beam bunching using code RON
• Measured phase versus end-gap change
• - Full range for one end 0.100 mm is
  0.16 period 29

17
Undulator End Phasing (contd) Complex Amplitude
of Radiation

• Complex amplitude of radiation, A(L) defines
  the intensity of radiation and is almost 100
  compared to ideal case
• Ideal case
• Regular part of the device cosine-type field
  distribution versus z
• Ends from measured data
• Measured slippage length for 113 periods of phase
  was 3.668 m at 6.35 mm gap (K value of 3.729)
• With two devices in a row, an error of 7x10-4 in
  ?Beff/Beff of the second device could be
  corrected by applying an end-phase correction of
  24.5 from both ends of the device
• Complemented by detailed RON simulations (R.
  Dejus) using random uniform distribution of Keff
  an error of up to 10x10-4 could be compensated
  by end phase correction

18
Undulator End Phasing (contd) Complex Amplitude
of Radiation

• Absolute value of complex amplitude of radiation
  versus z for two devices
• Second device field changed by 7x10-4
• Phase correction of 24.5 applied

Correcting phase of upstream end of 2nd undulator
is important for maximizing length of vector
(absolute value of radiation amplitude)
19
Undulator End Phasing (contd) RON Simulations

• ?Beff/Beff variation from undulator to undulator
  for 33 undulators with and without end-phase
  corrections _at_ 1.2 mm-mrad and 1.5 ?

Curves from top to bottom 1. Ideal case 2-4
with corrections 5-7 no corrections
20
Conclusions

• Measurements and tuning were done
• The prototype met all stringent mechanical and
  magnetic tolerances after a few design changes
  and magnetic tuning
• Tuning time is about two days after the exact
  effective field is set
• Setting of exact effective field with accuracy
  better than 1.5x10-4 is a challenge
• End-phase corrections of 29 total range allows
  compensation of ?Beff/ Beff of 8.2x10-4 or
  1.5C
• Lessons learned to simplify production
• - The biggest source of errors is variation
  of pole heights on assembled device. The
  tolerances for 1st prototype were 0.05 mm
• Do not need to measure individual magnet blocks
  in the half-period fixture with Hall probe with
  such pole height errors
• Mechanical tolerances of 0.025mm for gap
  uniformity will facilitate the tuning