Stability Issues

1

• Stability Issues
• NSLS-II ASAC Meeting
• April 23, 2007
• S. Krinsky

2
Stability Task Force / Workshop April 18-20
http//www.bnl.gov/nsls2/workshops/Stability_Wshop
_4-18-07.asp
Visiting Committee M. Boge PSI J. Byrd LBL J.R.
Chen Taiwan Y. Dabin ESRF R. Hettel
(Chair) SLAC J. Jacob ESRF J. Maser APS R.
Mueller BESSY-II D. Shu APS J. Sidarous APS O.
Singh APS C. Steier LBL
3
Electron Beam Sizes and Divergences for Selected
NSLS-II Sources
4

• User Requirements
• In most cases studied so far, a stability
  criterion of 10 of the beam size and 10 of the
  beam opening angle is sufficient, with the
  exception of the horizontal position for a few
  techniques
• Review Committee?Beam size stability also
  critical
• A common theme which has been expressed is in
  stability of beam intensity delivered to the
  experiment, which affects signal-to-noise
  directly, and this explains why some cases
  require beam position stability of lt10 of the
  beam size
• A one size fits all approach may not work for
  everyone, and tighter stability for a particular
  experimental program may require local measures

5
Review Committee Comments on Stability Solutions

• Need cutting-edge technology in many systems on
  BL and in accelerator
• May need mechanical motion/position survey
  sensors at critical points from source to
  experiment and in accelerator ability to include
  sensors in feedback
• Need to mechanically model critical beam line
  set-ups (supports, modes, etc)
• Find a way to monitor I0 just upstream of sample
  for all critical systems normalization on
  sample-by-sample but there are limits to
  quality of I0 detector
• Recommend phase space acceptance analysis
  projected to source phase space
• Use telescope technology to maintain relative
  stability of components (e.g. D. Shu)
• Need instrumentation infrastructure to verify
  accelerator vs. beam line stability issues and to
  help achieve stability goals
• Committee strongly supports beam designers goal
  to consider source and beam line stability
  holistically

6
Stability Dependent on Conventional Facilities

• Stability goals driven by conventional facility
  design
• Stability of storage ring tunnel floor
• Vibration lt 25 nm PSD from 4-50hz
• Stability of experimental floor
• Vibration level of lt 25 nm PSD from 4-50hz for
  general floor area
• Vibration level for 1 nm resolution beam lines
  requires further definition but appears
  achievable with proper correlation
• Thermal stability of storage ring tunnel
  environment
• /- 0.1o C for 1 hour time constant
• Thermal stability of experimental floor
• /- 0.5o C for 1 hour time constant
• Review Committee Accelerator group must confirm
  that there is no significant thermal
• load variation during operation

7
RMS (2 50 Hz) 20 nm
8
Ring Building Section
Bldg structure Isolated from tunnel and
experimental Floor
Isolation Joint or Void Space
Electrical Mezzanine
Isolated Grade Beam
Tunnel Roof
Ratchet or Shield Wall
Isolation Joint
Earth Shield Berm
Access Corridor
Experimental Floor
Monolithic Joint
Isolated Pier for Column
Tunnel Floor
9
Tunnel Design - Ring Building Section
Need to assure that vibration mitigation measures
are carried out at Ring building interfaces with
other structures and where systems enter building
or tunnel
Non-vibrating Equipment
Rotating Machinery
Non-vibrating Equipment
Rotating Machinery
Distance determined by modeling empirical
analysis
Section at Lab Office Building and Service
Building
10
Vibration Analysis

• Finite element calculations used
• to analyze effect of vibrations on facility

30 ft
11
Review Committee Revisit the project design
parameters regarding the infield service
buildings. From vibration prospective, it may be
better to locate them in the outfield (maybe
incorporated into LOBs) A discussion took
place, and CFG will pursue that approach from
cost/benefit approach. In either case, even
with the analysis resulting in acceptable
outcome, an attempt should be made to locate
rotating equipments as far away from SR as
practically feasible.
12
Mode Shapes of the Girder-Magnets Assembly
Review Committee Resonant frequencies often
found to be 1.5-2 times lower than calculation.
Must prototype magnet-girder assembly
13
Tolerances on Magnets Motion

• ?X Tolerance limits are easily achievable.
• ?Y Tolerance limits
• Thermal relative thermal displacement between
  magnets on the same girder lt 0.025 µm. (RMS
  thermal displacement of girders over a pentant (6
  cells) lt 0.1 µm)
• Vibration no magnification of ambient floor
  motion up to 50 Hz.
• Below 4 Hz girder motions are highly correlated
• Above 50 Hz the rms floor motion is lt 0.001 µm

14
Location of BPMs and
Correctors BPMs mounted on vacuum chambers
0.2 µm (vertical) User BPMs (upstream and
downstream of IDs) 0.1 µm (vertical) X-BPMs
0.1 µm (vertical)
There are also fast correctors in straights at
both ends of ID
Review Committee Include feed-forward on skew
quads to correct for ID changes
15
Support of Beam Position Monitors

• BPMs on the vacuum chambers need to be located
  near the fixed or flexible supports.
• Thermally insulated, sand-filled steel stands
  will meet the mechanical stability requirements
  for the special BPMs.

Review Committee Temperature of insulated
supports can change significantly over long
shutdowns. Must include method to quickly bring
supports to proper temperature at beginning of
new run.
16
Effect of Feedback
Without feedback
With feedback

• If G is a large positive number, with feedback
  loops on, the error signal is reduced by a
  factor of 1G at DC (T(0)1).
• At higher frequency, TG is a complex number and
  has to be designed to avoid oscillation.

17
Calculations show that 4 BPMs and 4 Correctors
per cell is sufficient to meet requirements.
Review Committee Consider global system Local
systems probably not needed Include maximum
flexibility in design for use of correctors and
BPMs Make provision for inclusion of x-ray BPMs
(2 per ID beamline) There are several proven
approaches to incorporating slow and fast
correction Digital technology is improving and 15
KHz data rate should be available Carry out real
time modeling of feedback system, include errors
in response Matrix.
18
Corrector magnets power supplies

• The system will use 120 corrector magnets with
  separate horizontal
• and vertical coils.
• The magnets will be designed for fast correction
  of 100 Hz.
• The dc transfer function of the magnet is 1000
  µrad per 19.2 Amps.
• The magnets will be located over stainless steel
  bellows and or
• flanges.
• The magnets are placed at the ends of each main
  dipole magnet.
• There will be 120 horizontal and 120 vertical
  power supplies.
• These corrector magnets and power supplies will
  be also used in
• slow and alignment corrections.
• The power supplies will have a high current
  requirement for
• slow/alignment corrections and high voltage
  requirements for fast
• corrections.

19
Corrector magnets power supplies

• The power supply requirements from accelerator
  physics are the following
• Frequency Strength - RMS
• lt 5 Hz 800 µrad
• 20 Hz 100 µrad
• 100 Hz 10 µrad
• 1000 Hz 1 µrad
• Resolution of last bit 0.01 µrad
• Noise Level 0.003 urad ( 4 ppm of 800 µrad
  ) (Review Committee ?1ppm)
• (These rating are for vertical correction and
  the horizontal correction is less stringent and
  they need to be quantified.)
• Power Supply Description
• Four quadrant switch-mode class D amplifier to be
  incorporated into a bipolar current regulated
  power supply .
• Small signal bandwidth of the power supply will
  be 2 kHz
• Amplifier has a switching frequency of 81 Hz.
  which gives a ripple current of 2 ppm.
• Resolution of 0.01 µrad is planed. Two 16 bit
  DACs will be used. One will be used for the slow
  large strength correction and the other for the
  fast smaller strength correction.
• Two DACs will have an affective resolution of 18
  bits or 0.003 µrad.

(Review Committee?20bits)
20
Main dipole power supply

• The power supply is a unipolar, 2-quadrant,
  current-regulated supply. It will use two
  12-pulse SCR converters in series with the center
  point connected to ground.
• Each converter will have a two-stage LCRL passive
  filter and a series pass active filter.
• Each main dipole magnet bending angle of 0.1047
  rad. The CDR has the current ripple spec.
• ( referred to Imax) of 5 ppm for freq. 60 Hz and
  greater. This gives a 524 nrad noise in the
  horizontal direction.
• CDR has the following power supply parameters
• resolution of reference current
  18 bit 1LSB
• stability (8 h-10 s) referred to Imax 40
  ppm
• stability (10s-300 ms) referred to Imax 20
  ppm
• stability (300 ms- 0 ms) referred to Imax
  10 ppm
• absolute accuracy referred to Imax 100 ppm
• reproducibility long term referred to Imax
  50 ppm
• To ensure long-term stability and reproducibility
  - high-precision DMMs will be used to monitor the
  power supply current, a redundant current sensor,
  and the analog current set point.
• RD for the main dipole ps is to develop a more
  thorough electrical circuit model of the system,
  that will include transmission line effects of
  the overall circuit.

21
Multipole power supplies for quad. sext. Magnets

• There is one power supply for each magnet.
• The power supply is a unipolar, single-quadrant,
  current-regulated switch-mode design.
• The power supply will use a DCCT as the current
  feedback device.
• To minimize current ripple, an additional output
  filter will be used.
• The CDR has the current ripple spec. ( referred
  to Imax ) of 15 ppm for freq. 60 Hz and greater.
• CDR has the following
• resolution of reference current
  16 bit 1LSB
• stability (8 h-10 s) referred to Imax 200 ppm
• stability (10s-300 ms) referred to Imax 200
  ppm
• stability (300 ms- 0 ms) referred to Imax 100
  ppm
• absolute accuracy referred to Imax 200 ppm
• reproducibility long term referred to
  Imax 100 ppm
• To ensure long-term stability and reproducibility
  - high-precision DMMs will be used to monitor the
  power supply current, a redundant current sensor,
  and the analog current set point.
• The RD for the multipole power supplies is to
  build a proto-type and confirm the accuracy,
  stability, and current ripple of the power supply.

(Review Committee?100ppm, 18bit)
22
RF BPMs

• Design similar to one adopted at RHIC
• 5-mm radius buttons
• Stray capacitance 1-4 pF (2p500MHz50O3pF0.5)
• Signal level -1.1 dBm for 500 mA at 500 MHz
• Dependence of vacuum chamber shape/size and
  button capacitance (and hence sensitivity) on
  fill pattern and circulating current can be
  significant

23
Processing Units

• Utilized at Elettra, NSSRC, Diamond, Soleil, PLS
• Fast acquisition 10 kHz sampling rate, 2 kHz BW
• Slow acquisition 10 Hz sampling rate, 4 Hz BW
• 32 bit data
• RMS uncertainty (for 10 mm scale in 1 kHz BW)
  -90.5dB ?0.3µm _at_ Pin -20 dBm
• 8-hour stability (?T1C) -80dB?1µm
• Temperature drift (T1035C) -94dB/C ? 0.2µm/C
  
• MTBF 100,000 hours
• For 270 units failure rate will be one unit in 17
  days

Review committee NSLS-II needs about factor of
2 better performance than available today
noise, stability lt0.15micron Technology
improving, in a few years will be achievable
24
Photon Beam Position Monitors

• Will provide information on photon beam position
  and angle (to account for errors in the wiggler
  field)
• Use of photon BPMs will allow sub-microradian
  pointing stability
• Contamination with dipole radiation can be of
  less concern due to reduced magnetic field in the
  bending magnet
• Can be used for orbit feedback and/or control of
  users optics
• 2D translation stages will precisely locate the
  photon BPM
• Should withstand high power density

Review Committee X-ray BPMs will be essential
for NSLS-II Give serious consideration to Decker
distortion Hold Workshop on X-Ray BPM Development