NSLSII Beam Position Monitor System

1
NSLS-II Beam Position Monitor System
Om Singh Instrumentation Group Leader ASAC Review
- October 22-23, 2009
2
Outline

• Injector RF BPM Button
• Storage Ring RF BPM Button
• RF BPM Electronics
• Photon BPM

3
Injector RF BPM buttons

• Resolution Requirement
• High charge (15nC) ? 30 microns
• Low charge (0.05 nC) ? 300 microns

• Simulation shows15 mm diameter buttons
• will meet resolution requirement
• BPM electronics - commercial or in-house

Pinayev Padrazo
Multi-bunch mode with 0.1 nC per bunch ? 15
ps Cb 2.5 pF
4
4
Injector RF BPM buttons - continued

• Sensitivity and power level calculations
  completed
• 15 mm dia RF buttons assembly design in progress
• Completion schedule 1st Article 6/2010
  Production 6/2011

Pinayev Padrazo Kosciuk
40x90 mm LTB beampipe 40x40 mm LTB beampipe
5
SR RF BPMs Location in a Cell
2
1
2
3
1
3
4
6
5
6
RF Cable Junction Box SR Tunnel

• Junction box to interconnect
• and to house passive components
• SiO2 cables used _at_ buttons
• LMR-240 used from box to rack

DellaPenna Kosciuk
7
SR RF BPM Button Design Trapped Mode
Optimization
NSLS-II button geometry ? 7 mm dia 16mm hor sep
Bunch Gaussian BW (?15psec)
Freq of trapped mode (r3.5mm)
Relative voltage induced
I. Pinayev, A. Blednykh, P. Cameron, B. Bacha
8
SR RF BPM Button Design Heating Optimization
FEA Thermal Simulation
1st Article Unit
Kosciuk Cameron
9
SR RF BPM Buttons Procurement Status

• Acquisition Status for large aperture (BLA)
  buttons (7mm dia)
• First articles received and acceptance tests
  completed ? August 09
• Contract for 460 pieces awarded ? Sep 2009
• Production schedule ? Dec 2009 June 2010
• Acquisition Status for small aperture (BSA)
  buttons (4mm dia)
• Engineering drawing, Specification, and SOW
  complete Mar 2010
• Contract Awarded - Apr 2010 Receive first
  articles Jul 2010
• Acceptance testing, go-ahead to vendor for
  production Sep 2010
• Order complete Dec 2010

P. Cameron
10
Resonance Modes in Multipole Chamber (3.05m)
Resonance modes can reside inside or too close
to BPM pass band frequency not good
Measurement Set-up
Blednykh Hseuh Ferreira Bacha
440
640
BPM Frequency Pass Band 492 508
MHz
Resonance Modes could effect BPM performance
and requires to be suppressed or moved out of
BPM band
RF Shields
RF Shields

• Use RF shield installed with modified NEG
  carriers in the ante-chamber
• Shield located 147 mm from beam center away
  from SR
• Shield 500 mm long with 300 mm spacing shifts
  Fo to gt 530 MHz
• Working on robust installation scheme Review
  committee - 9/25/2009

11
RF BPM ELECTRONICS

• 2/2009 Completed evaluation of Libera, Bergoz
  and APS BPM electronics. Libera meets NSLS-II
  baseline technical requirements
• 4/2009 - Shifted to evaluate other requirements -
  exposed several issues
• Utilizes 10 year old Virtex-2 FPGA technology
  - Virtex-6 is current
• Insufficient capability for future application
  software No room to grow
• Cumbersome to manage future upgrades FPGA
  source not given to NSLS-II
• Limited Itech support for a quick response at
  Brookhaven Located in Slovenia
• AFE Crossbar switching improves drift
  compensation, but adds complications for TBT and
  fast orbit feedback data.
• Expensive - 13,500 each (NSLS-II qty 250)
• 6/2009 Attended Libera User Workshop at ESRF
  mixed responses
• 7/2009 Motivated to evaluate an in-house BPM
  development production
• 10/2009 Planning design in advanced stage
• 12/2009 Make or buy decision finalized

12
NSLS-II RF BPM In-house Design Plan 7/2009

• Objective
• Design and build in-house RF BPMs for
• the Injection system and Storage Ring
• Development Approach
• Multi-phase development
• Parallel development of AFE and DFE
• System Integration Testing and Analysis
• Beam Test
• Iterate for performance optimization
• Dedicated Lab

• Challenges
• Aggressive schedule
• SR Stability Requirements
• SR Resolution Requirements

• BPM Development Team
• K. Vetter
• A. DellaPenna
• J. DeLong
• K. Ha
• Y. Hu
• B. Kosciuk
• J. Mead BNL Instrumentation
• S. Orban
• I. Pinayev
• Y. Tian

13
RF BPM Prototype - Physical Architecture (7/2009)
High Speed, Impedance Controlled Differential
Connector (LVDS)
ADC
BPF
Fixed Point DSP Cal
Control Interface
Raw Data
FOFB
A
First Turn
T-B-T (10KHz)
ADC
BPF
B
Post Mortem
Slow Data Control
Raw Data
Slow Acquisition
ADC
BPF
C
Timing
InterLock
InterLock
ADC
BPF
Virtex-5
D
Calibration
ADC Clock
Memory DDR2 SDRAM
AFE
DFE
AFE - Under Sampling (initial proposed
concept) - Factory Calibration (gain,
temperature) - Dynamic Calibration

• DFE
• - Fixed Point Digital Signal Processing,
  Position calculation, PU Linearization
• - Embedded Eventlink
• - Digitally Assisted AFE Calibration
• Factory Calibration Remove Systematic Errors
• Dynamic Calibration Remove Drift
• - Communication Links

14
Spectrum with Calibration

• Duplexer
• Frequency multiplex RF Signal from pickup with
  continuous calibration tone.
• Cal tone tracks RF Signal passband Insertion
  Loss by virtue of Monolithic Distributed design
• Currently working with vendors to optimize
  filter response, cost, and complexity
• Bessel/Gaussian 5th-order Transitional filter
  (e.g Gaussian to -3/-6dB, Chebychev) under
  investigation for RF Signal passband
• Bessel, Butterworth and Chebychev response under
  investigation for Pilot-Tone passband

RF Duplexer Theoretical Response Target
Rejection 60dB
Cal Tone Passband
RF Signal Passband
K. Vetter
15
Calibration Concept 1 - (Out-of-Band Pilot Tone
via RF Duplexer)
In-Tunnel hardware

• Benefits
• Continuous Real-Time Calibration
• No perturbations to signal path
• Inherent Built-In Test

K. Vetter
0.0007dB Gain Balance Required for 200nm offset
_at_ k10mm
16
RF BPM Prototype Development Plan (7/2009)
AFE
DFE
Beam Test
17
RF BPM Development Status 10/2009

• Started RF BPM development July 2009
• RF BPM architecture in mature stage (next slide)
• DFE Board Schematic 90 complete. Board layout to
  begin this month
• 1-month ahead of schedule
• Developing Requirements for custom RF Duplexer -
  Complete in November 2009
• ADC Evaluation started. Evaluating performance of
  four Candidate ADCs
• RF BPM Thermal Analysis started
• Highly desirable to have passive cooling (no
  fans)
• Take advantage of the temperature control rack
  capable of 0.1C stability
• Establish convective heat transfer coefficient
  inside the rack via accurate FEA study.
• With accurate boundary conditions and preliminary
  board layout, predict operating temperatures of
  FPGA and other electronic components.
• Start AFE layout in December 2009
• Preliminary chassis layout in advance stage
  more later
• Power supply requirements 85 complete
• RF BPM Development space Lab acquired Test
  equipment procurement in progress

K. Vetter
18
NSLS-II RF BPM Architecture 10/2009

• Bandpass Sampling
• 16b ADCs
• Sample Rate 113MHz
• Xilinx Virtex-5
• FOFB _at_ 10KHz
• SA _at_ 10Hz
• Embedded PC with Linux OS
• Embedded Floating-Pt coprocessor
• Embedded EventLink
• 128Mbyte DDR2
• Active Calibration for long-term drift
• Factory Calibration to remove systematic errors
• DFE platform adaptable to other NSLS-2
  applications

19
NSLS-II BPM Mechanical Concept
DFE AFE
Modular Approach - separate AFE, DFE
RF SMA Connectors
Triggers
Kosciuk Orban Tian
SFP Communication Ports
Same chassis as Power Supply Interface (PSI), PS
Regulator Module, and Cell Controller
20
NSLS-II RF BPM Electronics Schedule 10/2009
21
X-BPM Concept
Photo Emission Type XBPM assembly
B. Kosciuk P. Ilinski
Translation stages
Invar based High stability stand
22
Typical Front End Configuration XBPMs Shown
Ratchet Wall Collimator
Dual Safety Shutters
Shrama Doom
Fast Gate Valve
Photon Shutter
XBPM 1
XBPM 2
Lead Shield
Fixed Aperture Mask
Safety Shutter Collimator
Slow Gate Valve
Bending Magnet photon shutter
Ion Pump
X-Y Slits
Ion Pump with TSP
Ion Pumps with TSPs
Collimator
23
SUMMARY

• Injector RF button design in progress. Button
  delivery on schedule
• SR RF buttons
• Design completed including heating optimization
• 1st article accepted successfully Production
  delivery starts 12/2009
• Resonance modes solution in progress SR shield
  review held recommendations are being evaluated
• NSLS-II RF BPM Electronics
• Evaluated BPM electronics
• Libera meets technical requirements, but falls
  short in several other areas
• In-house design study in progress prototype test
  with beam 4/2010
• Installation schedule will be met with in-house
  BPM electronics

24
Acknowledgment
B. Bacha, A. Blednykh, A. Borrelli, P. Cameron,
W. Cheng, L.B. Dalesio, J. De Long, P. Ilinski,
A.J. Della Penna, L. Doom, M. Ferreira, G.
Ganetis, W. Guo, H-C Hseuh, E.D. Johnson, B.N.
Kosciuk, S.L. Kramer, S. Krinsky, F. Lincoln, C.
Longo, W. Louie, J. Mead, S. Orban, D. Padrazo,
I. Pinayev, J. Ricciardelli, G. Shen, S. Sharma,
J. Skaritka, C. Spataro, T. Tanabe, Y. Tian, K.
Vetter, W. Wilds, F.J. Willeke, L-H Yu
25
Back up slides
26
Rogue Mode Requirements - (From RF BPM
Perspective)
K. Vetter
27
9/2009 Libera Trip Report

• I-Tech has no plan to upgrade outdated HW (i.e.
  Virtex-2 or memory)
• Long-Term specification for Drift is 1um (NSLS-II
  Requirement 0.3um)
• 13KHz switching limited by GaAs switch settling
  time (i.e. Gate Lag Effect)
• Switch commutates at 13KHz and repeats at 13KHz/4
  3.3KHz. Periodic noise spikes exist at 3.3KHz
  and 13KHz. IIR filters are used to suppress
  transients
• Thermal failures observed for active fan RPM
  regulation
• Thermal failures fixed by limiting RPM range
  resulting in elevated operational temperature
  above target temperature by 15 deg.C

K. Vetter
28
Acceptance Testing of the 5 First Article Buttons

• Dimensional verification and leak checking done
  at the vendors facility.
• The following acceptance testing completed at BNL
• Vacuum Group - leak check, bake, leak check
  lt210-10 std.cc/sec
• Central Shops - dimensions verified using CMM
  (coordinate measuring machine)
• Diagnostics Group
• Capacitance 4pF, divided evenly between button
  and feedthru (one assembly shipped without
  buttons to permit this measurement)
• Contact insertion and withdrawal force
• Leakage resistance and voltage holdoff
• VSWR
• Trapped mode frequency (avoid the RF harmonics)
  and Q
• Thermal response to button heating
• In addition, Central Shops sectioned one assembly
  (wire EDM) to permit inspection of internal
  construction details.

29
Button Assembly 001, and Test Fixtures
Bacha Cameron Kosciuk
trapped mode frequency measurement
SMA contact insertion force measurement