(Microsoft PowerPoint - BIW10_\256\374\263\370)

∆
Amplitude
detector
3*RF
Σ
LP
Rocker I/O Interface
Heartbeat Register
Ethernet Interface
(Hardware UDP Stack)
Heartbeat Register
Control and Status
Registers
Horizontal Feedback
Vertical Feedback
Bunch Current Monitor
RF
Injection
Trigger
User
trigger
180
0
180
0
Slow Setting
Buffer
EPICS OPI
&
realtime and offline analysis
BIW10, TUPSM101, May 2-6, 2010
Design Status of the Diagnostic System for the TPS
K. T. Hsu, K. H. Hu, , C. H. Kuo, P.C. Chiu, Jenny Chen, C. Y. Wu, S.Y. Hsu
NSRRC, Hsinchu 30076, Taiwan
Abstract
Taiwan Photon Source (TPS) is a 3 GeV synchrotron light source which is being in construction at NSRRC. Designs of various
diagnostics are undergoing and will deploy in the future to satisfy stringent requirements of TPS for commissioning, top-up injection,
and operation. These designs which include beam intensity observation, trajectory and beam positions measurement, destructive
profile measurement, synchrotron radiation monitors, beam loss monitors, orbit and bunch-by-bunch feedbacks, filling pattern and
etc. are in final design phase. Details of current status and implementation of the planned beam instrumentation system for the TPS
will be summarized in this report.
Monitor
YAG:Ce screen
WCM
FCT
ICT
Monitor
FCT
Monitor
NPCT
FCT
Civil construction was started from February 2010.
The building will be finished in 2012.
Scheduled commissioning start in late 2013.
View form this direction
TLS
(existed 1.5 GeV light source)
YAG:Ce/OTR screen
Linac diagnostics
Quantity
5
1
2
1
Beam parameters
Position & Profile
Intensity distribution
Intensity distribution
Charge at exit of the linac
LTB diagnostics instruments
Booster synchrotron diagnostics instruments
BPM (4 button pick-ups)
Set of striplines and amplifiers
YAG:Ce screen (fluorescent
screen)
Synchrotron light monitor,
profile and streak camera
(visible light)
Bunch cleaning system
1
1
60
2
6
2
1
6
2
Fill pattern
-
TPS
TPS
7 m straight center
Dipole (1 o source point)
Introduction
The electron beam sizes and divergence
σx’
σy’
Source point
σx
σy
(μrad (μrad
(μm) (μm)
)
)
12 m straight center 165.10 12.49 9.85 1.63
120.81
39.73
17.26
76.11
15.81
Linac and Transport Line Diagnostics
Quantity Beam parameters
Quantity Beam parameters
Position, profile(1 at diagnostic branch). OTR screen
will be adopted for the site of high precision profile
measurement to avoid saturation of YAG:Ce screen.
Beam intensity
5.11
3.14
1.11
Circumference (m)
Energy (GeV)
Natural emittance (nm-rad)
Revolution period (ns)
Revolution frequency (kHz)
Radiofrequency (MHz)
Harmonic number
SR loss/turn, dipole (MeV)
Natural chromaticity ξx /ξy
Dipole bending radius ρ(m)
Repetition rate (Hz)
Major parameters of the TPS booster
synchrotron and the storage ring
Booster Synchrotron
150 MeV – 3 GeV
10.32 @ 3 GeV
603.865
499.654
0.586 @ 3 GeV
Betatron tune νx /νy
14.369/9.405 26.18 /13.28
Synchrotron tune ns
-
0.00611
Momentum compaction (α1 , α2 ) -
2.4×10 -4 , 2.1×10 -3
Natural energy spread
9.553×10 -4
8.86×10 -4
Damping partition J x /J y /J s 1.82/1.00/1.18 0.9977/1.0/ 2.0023
Damping time τx /τy/τs (ms) 9.34/ 16.96 / 14.32 12.20/ 12.17 / 6.08
-16.86/-13.29
17.1887
TPS 150 MeV linac has been contract to RI Research Instruments GmbH.
The schedule for delivery and commissioning is in early 2011 at test site.
The linac will move to the TPS building in late 2012 after TPS building available.
All diagnostics of the linac system is provided by the vendor.
YAG:Ce screen monitors for beam position and profile observation.
The fast current transformers (FCT) for monitor the distribution of charge.
The integrating current transformer (ICT) for monitoring total bunch train charge.
496.8
1656
828
BTS diagnostics instruments
3
Storage Ring
ICT
1 Beam charge
ICT
2 Beam charge, installed at upstream
BPM and single pass electronics 7 Beam position
and downstream of the BTS
Energy define slit
1 1 pair of horizontal jar
BPM and single pass electronics 6 Beam position, relative intensity
The YAG:Ce fluorescence screens will provide information of beam position and profile.
The OTR screens are also considered to be used for high precision of beam emittance and energy spread measurement
The ICT will provide information of beam charge pass LTB and BTS and hence the beam losses during the injection cycle.
The beam trajectory will be monitored with BPM equipped with Libera Brilliance Single-Pass, its functionality for single pass measurement.
Averaged beam current
Beam position
Betatron tune, bunch
cleaning system
Beam profile and position
at injection, extraction,
and at every lattice cells
Beam size (emittance),
bunch length
TLS
Monitor
YAG:Ce/OTR screen
FCT
Booster Ring Diagnostics
Quanti
ty
3
1
Beam parameters
Position, profile, booster
extraction beam emittance
Beam intensity
518.4
3.0
1.6
1729.2
578.30
499.654
864
0.85269
-75 / -26
8.40338
Fluorescent screens will be installed at injection and extraction section and at the other
lattice cells to facilitate booster commissioning, troubleshooting and psychology needing.
Booster orbit will be monitored with 60 BPMs with turn-by-turn capability.
The sum signal from the receivers can be used to monitor fast history of the beam current.
Circulating current will be measured with NPCT, while bunch pattern will be monitored
with FCT.
For tune measurement, the electron beam will be excited with white noise using striplines.
The beam response will be observed with a real-time spectrum analyzer connected to the
dedicated BPM buttons with the front end.
There will be an extra set of striplines for a bunch cleaning system and for users who need
a specific filling pattern in the storage ring.
Synchrotron radiation from a dipole will be used to observe the beam profile during
energy ramping and emittance measurements.
The capability to monitor bunch length with a streak-camera will be also provided.
-
Monitor
NPCT
Sum signal of BPM buttons
BPM (4 button pick-ups)
BPM (4 button pick-ups)
Striplines
Transverse kickers and
amplifiers
YAG:Ce screen
(Fluorescent screen)
PIN diode type beam loss
monitors
Scintillation loss monitor
Scrapers
Storage ring diagnostics instruments
Synchrotron radiation diagnostics instruments
Monitor
X-ray pinhole camera
Streak camera
Time correlated single
photon counting system
(Visible light or X-ray)
XBPM
Visible light synchrotron
light diagnostic station,
Imaging and streak
camera
BPM
TPS bunch-by-bunch feedback
system
Hybrid Network
bunch
Quantity
2
1
1
1 or 2 per
beamline
Possible cPCI DAC Configuration
10 kHz Fast Setting Clock
DO (LEMO connector, for timing measurement), Trigger out, Package received
4
Setting Buffer
Fast SFP
Setting #1 Port
Fast
Setting #4
2 Rocket I/O ports
2 GbE ports
cPCI Bus
Slow
Setting
Trigger
10 kHz ~ 250 MHz
FPGA
Class A Power
Amplifier x 4
Σ ∆ x ∆ y
Front-end
FPGA
x
GbE
Port
One of an Implementation Option
Distributed IFOB embedded in BPM modules
BPM Electronics
12bit ADC
12bit ADC
12bit ADC
FPGA
cPCI 32 ch, 20/18 bits
DAC Module
Slow Setting
EPICS IOC
Fast Setting
16bit DAC
16bit DAC
16bit DAC
Cell N Power Supply
BPM Electronics
Control Network
1
Beam parameters
Emittance vertical and horizontal
planes.
Averaged profile, single turn profile,
profile of the selected bunch.
Bunch length.
Cell N-1 Cell N+1
XBPM satellite grouping
BPM grouping and
XBPM
XBPM
IFOB FPGA
Electronics
Electronics
Timing Interface
Control and Status Registers
Cell N
Readback
Registers (6
x 32)
Quantity
4~6 per
cell
2 sets per
plane
+
1
1
168
1
1
2
1
24
XBPM
Data Sniffer
Filling pattern.
Isolated bunch purity.
Position and angle of ID radiations.
Alternative beam size measurement
(emittance), either imaging the
vertical polarized synchrotron light
or interferometer, bunch length.
20/18 bit DACs
Beam parameters
Averaged beam current, beam lifetime
Fill pattern, bunch current
7 BPM/cell
For bunch-by-bunch feedback pickups
Betatron tune measurement
Horizontal and vertical kicker for transverse
feedback and bunch cleaning usage.
Beam profile and position just after injection
septum
Beam loss pattern
High counting rate type beam loss monitor
1 set = 2 blades
Storage Ring Diagnostics
The storage ring diagnostics including averaged beam current, fill pattern,
beam lifetime, closed orbit, working tunes, chromaticity, beam size, beam loss
pattern, beam density distribution, emittance, bunch length, and etc.
The NPCT provides a resolution of better than 1 µA/Hz 1/2 and has large
dynamic range and bandwidth to make itself a versatile device for measuring
lifetime and injection efficiency.
The storage ring filling pattern observed from the sum signal of BPM buttons
by wide bandwidth fast digitizer sampling at RF or a multiple of RF frequency
will enable measurement of the bunch current to better than 0.5% accuracy.
Libera Brilliance BPM was chosen as baseline design at the conceptual design
phase.
Updated BPM electronics platform equip with more advanced parts and
enhanced functionality are in serious consideration and discussed with the
possible vendor. The TPS will most likely adopt the new BPM platform.
The photon diagnostics for the TPS storage ring will utilize visible and X-ray
synchrotron radiation generated in a bending magnet.
Visible light beamline will be built to measure various beam parameters by streak
camera, CCD camera and interferometer.
Synchroscan mode operate for the streak camera at 250 MHz is preferred. Integrating
the streak camera system with EPICS is preferred.
Two X-ray pinhole cameras imaging the electron beam from bending magnets is the
baseline design for the emittance measurement and measure the electron beam size at all
currents from < 1 mA to 400 mA.
The X-ray photon BPMs (XBPM) will be installed at each beamline. The slow data for
control system access and the fast data for feedback purpose.
Prototype of Libera Photon has been testing intensively at the 1.5 GeV TLS.
Measuring the filling pattern by using time correlated single photon counting (TCSPC)
is also considered. And APD detector to detect scattered X-ray photon will provide signal
input for the TCSPC system. More than six order of dynamic range are expected.
Infrastructure for orbit measurement, control and feedback
Slow orbit acquisition will perform by channel access to the BPM platform embedded EPICS IOC up to 10Hz
rate.
Fast orbit beam position will circulate around all BPM platform at 10 kHz rate by using FPGA grouping
scheme (e.g. Diamond Communication Controller or Gigabit Ethernet Grouping) or a new design.
The TPS will adopt special design high performance corrector power supply. The power supply will use analog
regulator, adopt bias analogue PWM scheme to improve zero current crossover problem.
Improves integrated noise level form DC to 1 kHz down to a few part per million of the output full scale
corresponding to nano-radian level kick for slow corrector with maximum ±600 µrad kick.
Control of each cell’s corrector will be through a special design 20 bits (or 18 bits) DAC module. This module
will provide EPICS CA interface via cPCI backplane for configuration, and setting and status monitoring.
The 10 KHz rate data stream fast setting ports might configure as Rocket I/O for orbit feedback application
and Gigabit Ethernet for global feed-forward applications.
Functionalities of this FPGA module will perform: one for BPM data grouping and the other for feedback
engine. BPM data grouping provides a way to distribute all BPM and XBPM data around the TPS storage ring at
all BPM platforms in 10 kHz rate.
Orbit feedback computation will distribute to the FPGA modules installed at the BPM platform in each cell.
Sniffer nodes will be setup to capture orbit information in 10 kHz rate for more than 10 sec and decimated data
at lower rate for much longer record time for various applications and analysis.
Bunch-by
by-bunch bunch feedbacks and diagnostics
Transverse coupled-bunch instability mainly caused by the resistive wall impedance and other sources will
deteriorate beam quality. Two plane Bunch-by-bunch feedback system is planned to suppressed instabilities.
Transverse feedback kickers are planned to adopt the SLS/ELETRA design and compatible with TPS vacuum
vessel. Transverse signals pick-up will be used as an extra BPM and installed at location of high beta function.
Beside feedback functionality, the feedback electronics and software also support bunch oscillation data capture
for analysis to deduce rich beam information, tune measurement, bunch clearing, and beam excitation and etc.
Features of the planned system include the latest high dynamic range ADC/DAC (12/16 bits), high performance
FPGA, flexible signal processing chains, flexible filter design, bunch feedback, tune measurement, bunch cleaning,
various beam excitation scheme, flexible connectivity, and seamless integrated with the control system.
On-line control interface to operate feedback system and off-line analysis tools should be included.
Testing of the Libera Bunch-by-Bunch and the iGp are on going at the 1.5 GeV Taiwan Light Source.
Summary
Beam diagnostics designs and implementation for the TPS are in proceed. Status is summarized in this report. The critical
diagnostic systems, addressing beam stability and low emittance monitoring, are being investigated in the design phase. Major
procurement are scheduled in 2011~2012. Optimizing the design, prototyping and working out on specifications are current efforts.
System integration is planned in 2013. Delivering a best diagnostics system to satisfy stringent requirements of TPS is the goals.

Control
30
20
10
0
-10
-20
20
15
10
5
0
-5
-10
Time(sec)
Time(sec)
RCVCSPS21 current (mA)
FOFB status
BL10y
r3bpm2y
r3bpm3y
20
10
0
-10
-100
-150
-200
-250
0
-200
-400
EPU4.6 Phase change (mm) v.s. photon BPM Ib = 232.8672 mA;
Time (sec)
BIW10, TUPSM102, May 2-6, 2010
Diagnostics Update of the Taiwan Light Source
C. H. Kuo, P. C. Chiu, Y. S. Cheng, Y. K. Chen, K. H. Hu, C. Y. Wu, Demi Lee, S. Y. Hsu, Jenny Chen, Y. T. Chang, C. J. Wang, Y. R. Pan, K.T. Hsu
NSRRC, Hsinchu 30076, Taiwan
Diagnostics of the 1.5 GeV Taiwan Light Source (TLS) has been continuously upgraded since 1993. The BPM electronics of the TLS
have been upgraded to the Libera Brilliance in August 2008 to improve performance and functionality. Orbit feedback system is also migrated
into fast orbit feedback system to enhance orbit stability. Commercial photon BPM electronics was tested recently. New generation bunch-bybunch
feedback processor was tested to improve beam stability. Post-mortem diagnostic tools were also set up to clarify reasons of beam trip.
These upgrades are contributed to improve beam quality and machine availability a lots. These efforts will be addressed in this report.
• The New BPM system has
been commissioning in August
2008. The long-term reliability
is satisfied during observation
for one and one-half years.
• The resolution of 10 kHz data
could be achieved 0.2 μm in
rms.
• Turn-by-turn data is provided
by Libera Brilliance. Post
mortem buffer with the length
up to 256k could store beam
turn by turn data before beam
trip.
• The commissioning of the new fast orbit feedback system
has started from 2008.
• The reflective memory is employed to shares fast orbit
data without consuming extra CPU resource.
• Libera Grouping is used to reduce Ethernet jitter.
• Pseudo-random binary sequence (PRBS) excitation is
employed to measure system response and latency and
then choosing the proper correctors for FOFB.
• The I/O latency time is around 500 msec. The computation
latency is about 120 msec.
• Singular value decomposition (SVD) is used to invert
response matrix. Tikhonov regularization is also adopted.
• The new FOFB is promoted to suppress noise of
bandwidth to more than 50 Hz from old 6 Hz system after
all components are upgraded. Increase loop bandwidth
will be done after further study.
Amplitude(µm/Hz 1/2 )
RMS Integrated PSD(µm 2 /Hz)
Hor. Position (µm)
Ver. Position (µm)
1.5
1
0.5
0
10 0
10 -2
10 -4
5
0
Abstract
New BPM System and Some Observation Delivery by the System
-300
-320
-340
-5
0 10 20 30 40 50 60 70 80 90 100
FOFB OFF
FOFB ON
5
0
-5
0 10 20 30 40 50 60 70 80 90 100
Time (sec)
FOFB OFF
FOFB ON
Hor. pos ition (µ m )
10 0 10 1
10 0 10 1
Frequency(Hz)
The overall RMS orbit stability could be
Hor. am plitude (µ m )
submicron from DC to 50 Hz.
• Some fast beam motion could be
observed by the new BPM system.
10
5
0 0.005 0.01 0.015 0.02 0.025 0.03
Time (sec)
0
0 100 200 300 400 500
Frequency (Hz)
The beam orbit motions at 120 Hz
caused by one problem corrector power
supply.
New Orbit Feedback System
y Position σ (µm)
y Position σ (µm)
1
0.8
0.6
0.4
0.2
WS/Unix
200 Hz Rate
Data
Acquisition
(200 sec 10 kHz
Diagnostic) Rate Data
Acquisition
(8 sec Diagnostic)
V
MPMC
E RM
V
H
MPMC
O
E RM
S
DI
T
H /
O DO
S
T
GbE Interface External/Post-mortem
Trigger
Libera Group R1 GbE Grouping
Libera Group R2
Libera Group R3
Libera Group R4
10 kHz
Rate
Data
Acquisition
V
MPMC
E RM
H
O
S
T
Libera Group R5
Libera Group R6
Super-period 1, 2, 3, 4, 5, 6 BPM
0
0 10 20 30 40 50 60
1
0.8
0.6
0.4
0.2
FOFB OFF
FOFB ON
0
0 10 20 30 40 50 60
BPM ID
Standard deviation of all BPM
SA reading (10 Hz rate) can be
reduced to 0.2 μm during
FOFB operation in the normal
user mode for the vertical
plane.
R1 BP M 1 H O R. D isp . (µm )
0
-100
-200
-300
-400
-500
0.02 0.04 0.06 0.08 0.1 0.12
Time (sec)
VME Host
PowerPC 7448 @ 1 GHz
PMC
RM
DI
Horizontal
Network
16
Bit
D
A
C
PC/Linux
Reflective Memory
Hub
VME Host
PowerPC 7448 @ 1 GHz
PMC
RM
Vertical
Control
Consoles
DI
16
Bit
D
A
C
Correction Magnet Power Supply
V
M
E
H
O
S
T
16
Bit
D
A
C
PC/Linux
16
Bit
D
A
C
V
MPMC
E RM
H
O
S
T
Plane
Plane
Inhibit Control Inhibit Control Corrector DC Control
GbE Interface
Fast Orbit Feedback setting (÷5 reduced VME
GbE
range)
Grouping
PS Control
Libera Phton
Analogue Sum
Libera Phton
250Injection
every 60 sec
Vertical Position (µm)
200
150
100
50
0
Before
After
Hor. injection amplitude (µm)
Ver. injection amplitude(µm)
1000
500
0
-500
-1000
0 10 20 30 40 50 60
200
100
0
-100
-200
0 10 20 30 40 50 60
BPM index
The orbit distortion caused by septum
leakage before and after chamber
shielding. The excursion is improved to
reduce around one half after chamber
shielding.
The infrastructure of the new orbit feedback system
VME
RM
ILC12
16
Bit
A
D
C
16
Bit
A
D
C
Vertical Position vs. Time; Ib = 302.03 mA; 07-Apr-2009 04:08:03
Future Option
Photon BPM
Data
Acquisition
XBPMs
EPU5.6 Phase Change (1 mm/sec)
10 kHz Rate Data
Acquisition
r1bpm0y
r1bpm1y
r1bpm2y
r1bpm3y
r1bpm4y
r1bpm5y
r1bpm6y
r1bpm7y
r1bpm8y
r1bpm9y
r2bpm0y
r2bpm1y
r2bpm2y
r2bpm3y
r2bpm4y
r2bpm5y
r2bpm6y
r2bpm7y
r2bpm8y
r2bpm9y
r3bpm0y
r3bpm1y
r3bpm2y
r3bpm3y
r3bpm4y
r3bpm4ay
r3bpm5ay
r3bpm5y
r3bpm6y
r3bpm7y
r3bpm8y
r4bpm0y
r4bpm1y
r4bpm2y
r4bpm3y
r4bpm4y
r4bpm5y
r4bpm6y
r4bpm7y
r4bpm8y
r4bpm9y
r5bpm0y
r5bpm1y
r5bpm2y
r5bpm3y
r5bpm4y
r5bpm5y
r5bpm6y
r5bpm7y
r5bpm8y
r5bpm9y
r6bpm0y
r6bpm1y
r6bpm2y
r6bpm3y
r6bpm4y
r6bpm5y
r6bpm6y
r6bpm7y
0 50 100 150 200 250 300 350
Time (sec)
Feedback ON Feedback OFF Feedback ON
Effect of the FOFB to suppress orbit excursion
due to phase change of EPU5.6
Before
After
Before
After
• There are several kinds of
photon BPMs with different
electronics and data acquisition
installed at beamline front-ends
at the TLS. To provide a better
integration and efficient usage
of the photon BPMs, integrated
solution is in study. Evaluate of
the Libera Photon is on going.
• Excellent results was achieved.
Seamless integration with
control system are foreseeable.
Feedback Test
Bunch Cleaning Test
• iGp has integrated functionality
for bunch cleaning.
• Apply excitation at the vertical
betatron tune.
• The right figures show two
cleaned out bunches patterns.
• Killed all bunches except for one
bunch, used for timing
optimization.
V e rt ic a l p o s itio n (µ m )
30
25
20
15
10
5
0
-5
0.62 0.64 0.66 0.68 0.7 0.72 0.74 0.76 0.78 0.8 0.82
Time (sec)
Libera Photon Test
BL10 Y
Injection transient observed
by the photon BPM (10 kHz
rate data).
Current (mA)
Vertical position (µm)
-30
0 10 20 30 40 50 60
FOFB OFF
FOFB ON
-15
0 10 20 30 40 50 60
iGp Bunch-by
by-Bunch Signal Processor Commissioning
• Based on existed bunch-by-bunch front-end, back-end, and
power amplifiers.
• Three planes was tested.
• Longitudinal feedback test – Beam stabilized in 20 minutes.
• Horizontal and vertical planes take a coupled of hours to make
the system work.
• 16-tap FIR filter.
• One loop to stabilized instabilities in horizontal and vertical
plane was also performed by using two peaks filter and got
success results.
• Matlab filter generator allows independent control of gain and
phase for X and Y.
• Lower X gain is intentional - to balance the two loops.
• To identify the reasons of trips at TLS,
various diagnostics tools are employed.
• An earthquake detector has been
installed to reveal the consequence of the
trip caused by earthquake.
• The right figure shows that an
earthquake occurred at high sea of
eastern Taiwan. The P wave arrive
NSRRC site 42 seconds later. The S
wave lag about 30 seconds after P wave.
• The SRF system trip happened 20
second after S wave arrived. The trip
was caused by the vibration of Nb cavity
which resulted in deformation of the
cavity and consequently provoked
resonance frequency shift.
Cleaning is in proceed
Post-mortem Diagnostics
Beamline #10 photon BPM 10 Hz
reading signal related to the
electron BPM when driving
corrector with/without FOFB.
Grow/Damp Test
-20
0 50 100 150 200 250
FOFB OFF FOFB ON
• Grow/damp transient at 200 mA is measured.
• Damping rate is twice as fast as the growth rate.
• Damping rate is twice as fast as the growth rate at least for
all three planes.
Horizontal plane
Most of the bunches killed
Sequence of the earthquake and beam trip
Phase change (mm)
BL10 ver. pos. (µm)
BL15 ver. pos. (µm)
-300
0 50 100 150 200 250
-600
0 50 100 150 200 250
Photon beam displacement
when the EPU5.6 phase
changed with/without FOFB.
Beam excursion could
almost be suppressed.
Vertical plane
Only the selected bunch survive
Summary
Longitudinal plane
• Diagnostics of the TLS has been
continuously upgraded using
available resources to improve
performance and avoid obsolete.
• New BPM system and fast orbit
feedback system are commissioning.
Performance tests are presented.
• Photon BPM are testing.
• Testing of the iGp and Libera
Bunch- by-Bunch are ongoing.
• Other diagnostic tools will be
continuously added and built up to
enhance system’s reliability and
availability.