new master oscillator at flash
TRANSCRIPT
New Master Oscillator at FLASH
Krzysztof Czuba *, Henning Weddig #
*Institute of Electronic Systems, Warsaw, [email protected]
#DESY, [email protected]
FLASH Seminar, 7 Oct. 2008
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 2
People Involved in the MO Project
Henning Weddig: requirements, concept, schematics, contacts with industry
Krzysztof Czuba: hardware and measurement work, diagnostics, planning and organization of students work, documentation
Frank Ludwig: helped managing the installation in June/July 2008
Thomas Froehlich: hardware installation in FLASH
Erhard Salow (Inwave Company): design and manufacturing sub-assemblies
Students (assembly, measurements, experiments): Bastian Lorbeer, Andrzej Stefanski, Michal Ładno, Jan Piekarski, Dominik Sikora, Łukasz Zembala
Technicians: Bartlomiej Szczepanski, Bibiane Wendland
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 3
Requirements for the MO System
• Generated and distributed frequencies: 50 Hz, 1 MHz, 9 MHz, 13.5 MHz, 27 MHz, 81 MHz, 108 MHz, 1.3 GHz, 1.517 GHz, 2.856 GHz
• Required signal phase stability (@ 1.3 GHz):– Short term: 0.1 ps (integrated timing jitter 10 Hz – 1 MHz)– Long term (1000s): 1.0 ps– Long term (days): 5.0 ps
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 4
Basic Decisions
- OCXO reference oscillator
- RF frequency synthesizer and frequency divider network
- 19” crate system
- Diagnostics incorporated into MO crates
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 5
Some History ...
● In 2001 a 20 page requirement document created by H.W. was sent to industry. Only one company offered building the MO. Others claimed they could not do it or offered very expensive research
● Requests for new signals, frequencies and powers, more tap points forced changing concept and rebuilding ready boxes. This influenced the growth of the diagnostic system and the power supply.
● Many problems had to be solved: e.g. cabling, power supply filtering, ground loops, heat dissipation, sensitivity to mechanical vibrations, parts breaking down (manufacturer error), commercial components not fulfilling specs....
Very first New MO1 crate!
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 6
Basic System Conception● Six 19” crates
● Diagnostics integrated with each crate
1.3 GHz PLL 1.3 GHz POWER AMPLIFIER
1.3 GHz
PA – Power Amplifier
44 dBm
30 dBm
LOW POWER PARTOCXO
9 MHz
~ 81 MHz PLLx 9
108 MHz PLLx 12
Div. : 3
Div. : 2
Div. : 9
Div. : 9
PLL x 144
PA
27 MHz
13.5 MHz
9 MHz
1 MHz
108 MHz
PA
1.3 GHz DISTRI-BUTION BOX
81 MHz POWER AMPLIFIER
81 MHz40 dBm
PA
81 MHz DISTRI-BUTION BOX
DDS 50 Hz
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 7
Low Power Part – Crate Design Example
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 8
Low Power Part – Assembly
IN
OU
T
OU
T
SPLITTER
SM
A
SMASMA
OUT
9 MHz PA
SMA
SM
A
SM
A
SM
A
SM
A
SM
A
INO
UT
9 MH
z TT
L DR
IVE
R
SM
AS
MA
INO
UT
1 MH
z T
TL DR
IVE
R
SM
AS
MA
5V POWER SUPPLY
DIV
IDE
R 9/1
IND
ET
EC
L O
UT
SIN OUT
SM
A
SM
A
SMA
SMA
SMA
DIV
IDE
R 81/9
IND
ET
EC
L O
UT
SIN OUT
SM
A
SM
A
SMA
SMA
SMA
DIV
IDE
R 27/13.5
IND
ET
EC
L O
UT
SIN OUT
SM
A
SM
A
SMA
SMA
SMA
DIVIDER 81/27
IN DET
ECL OUT
SIN
OU
T SMA
SMA
SM
A
SM
A
SM
A
SMA
SMA
LOC
KM
ON
ITO
R
SMA
SMA
LOC
KM
ON
ITO
RC
PL
INOU
T
COUPLER
SM
A
SMASMA
PLL
5S
MA
SM
A
SM
A
RE
FFI
N
Vtu
nePLL
5S
MA
SM
A
SM
A
RE
FFI
N
Vtu
ne
IN
OU
T
OU
T
SPLITTER
SM
A
SMASMA
SMA Vtune
OUT
81 MHzVCXO - PA
SMASMA Coupl
SM
A
SM
A
SM
A
SM
A
50 Hz
DD
SD
DS
50Hz
SMA Vtune
OUT
108 MHzVCXO - PA
SMASMA Coupl
SM
A
SM
A
SM
A
SM
A
CP
L
IN
OU
T
COUPLER
SM
A
SMA SMA
PO
WE
RD
ET
EC
TOR
OUT
IN
SM
A
Vcc
CP
L
INOU
T
COUPLER
SM
A
SMASMA
SMA
9MHzOCXO
12V POWER SUPPLY
FRONT
Power
9 MHz
108 MHz Lock
1 MHz
9 MHz
9 MHz OCXO 9 MHz PA
27 MHz
81 MHz VCXOPA
108 MHz VCXOPA 13.5 MHz
Signal Amplitude
Ext. Control In.
MASTER OSCILLATOR LOW POWER PART
+20 dBm1MHz TTL
+20 dBm +20 dBm +20 dBm9 MHz TTL 13.5 MHz
+8 dBm27 MHz+11 dBm
13.5 MHz+8 dBm
27 MHz+11 dBm
1MHz+8 dBm
108 MHz+20 dBm +20 dBm+20 dBm +20 dBm
108 MHz Monitor Out
50 Hz
Control 81 MHz Lock
Supply Voltage 12VSupply Voltage 5V
Int. Ext.
81 MHz Monitor Out
81 MHz+20 dBm +20 dBm+20 dBm +20 dBm
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 9
Distribution Box, HPA, PLL
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 10
Power Supply● Linear regulated power supply
● Battery backed up power supply for LPP and 1300 MHz PLL
● Unbreakable operation due to battery allows for maintaining signal phases in case of short mains power failure
● Diagnostics with Ethernet Interface
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 11
Before Installation ...
Finding proper connectors was a real challenge (no labels, many blind paths, etc...)
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 12
System Assembly - Concept
● Redundancy (two MO copies)
● Short cable (jumper) based switch
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 13
Installed MO System Scheme
MO RACK 1 MO RACK 2 MO RACK 3
RACK 1 PATCH PANEL TOP RACK 2 PATCH PANEL TOP RACK 3 PATCH PANEL TOP
LPP
PLL1300 PSI
RF CablesRack 1Page 3 RF Cables
Rack 3, page 5SWITCH PANEL
1.3GHz DISTRIBUTION BOX
81MHz DISTRIBUTION BOX
BOTTOM
TOP
BOTTOM
TOP
BOTTOM
TOP
RF Cables from Rack 1 to Rack 2,
page 4
PA
TC
H P
AN
EL
- C
AB
LES
RF Cables from Rack 3 to Rack 2, page 6
RF CablesRack 2, page 7
DistributionCables
RF Cables Rack 2, page 8
TABLE 3 TABLE 4
TABLE 5 TABLE6
TABLE 7
24 24
24
24 24
TABLE 1
TABLE 2
108MHz DISTRIBUTION BOX
9MHz DISTRIBUTION BOX
TABLE 8
TABLE 9
TABLE 10
TABLE 11
FREQ. COMPARISON BOX
PLL1517
HPA1300
HPA81
AQB (Data Acquisition)
AQB Patch PanelAQB (Data Acquisition)
AQB Patch Panel
LPP
PLL1300 PSI
PLL1517
HPA1300
HPA81
AQB (Data Acquisition)
AQB Patch Panel
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 14
Installed MO System
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 15
Rack 1 and Rack 2
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 16
MO Distribution Cables in FLASH
108MHz
10m
TTF CONTROLROOM
LLRF control racks
LOLA
165m 85m 19m140m260m 118m 105m 93m 56m 10m 0m44m
waveguide
20m
New LLRF control racks
18m
Klystron 3
RF GUNACC1ACC2ACC3ACC4
ACCELERATOR TUNNEL
HALL 3 extension HALL 3
EOS
RackLINAC 15 LINAC 9
Rack
LASER
ACC5ACC7 / ACC6LOLA
95m 40m56m
Klystron 2Klystron 1
Klystron 5
Klystron 4
MASTEROSCILLATOR
1.3GHz
1.3GHz
2856MHz
9MHz
13,5MHz
Cable 1
Cable 2
Cable 4
Cable 3
Cable 5Cable 15
Cable 6
Cable 10
Cable 11Cable 12
Cable 147/8" Cable 7
Cable 9
Cable 13Cable 8
Cable 22 1517MHz
1.3GHz
Fiber-Optic Cables
Cable 19Cable 20
Cable 21
1.3GHz
9MHz
81MHz7/8 " cable
7/8 " cable
1/2 " cable
27MHz
InjectorArea
EXPERIMENTAL
HALL
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 17
Diagnostics
Measured are:● Power levels● Phase changes over amplifiers● VSWR of most important cables● Crate temperatures● PLL lock status● VCO control voltages● Power supply voltages and currents
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 18
Data Acquisition for Diagnostics
PS Eth
Master Slave 0
Slave 1
Blind Blind Blind Blind
BLIND
M D M Ain S0 AinS0 D S1 AinS1 D S2 AinS2 D S3 AinS3 D S4 AinS4 D S5 AinS5 D S6 AinS6 D
4 TE
Analog Patch Cable
Digital Patch Cable
AQB CRATE
AQB PATCH PANEL
Slave 2
Data Acquisition PCB
Monitorig Crate and Patch Panel
PCB design and crate assembly by Michał Ładno
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 19
Diagnostics Software
PCB Firmware:
● Developed for the needs of the project
● Allows for data acquisition and communication within crate
● Allows for system extending without a need for PCB reprogramming
Server software features:
● Communication with crates
● On-line data acquisition and storage
● Software configuration of channel parameters and storage resolution – large flexibility
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 20
MO DOOCS Panels
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 21
LPP DOOCS Panel
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 22
Power Supply Panel
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 23
Frequency Comparison Box Panel
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 24
Laboratory Tests
● All MO crates were tested in laboratory
● Detailed test protocols available
Jitter ~36.7 fs
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 25
Measurements in Hall 3
● Phase noise levels measured in Hall 3 do not differ significantly from lab. measurement results!
● There is still need to investigate phase drifts
1.3 GHz phase noise in Hall 3
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 26
Current MO System Status
● Rack 1 and Rack 2 completed● Power supply for Rack 1 installed● Rack 3 crates ready in the lab, not installed yet● Power supplies for Rack 3 missing (ordering in
progress)● Diagnostic system final debugging in progress
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 27
Future Plans
● Install crates in Rack 3 and power supplies● Install Frequency Comparison Boxes in Hall 3 Ext. and
in Experimental Hall – important for drift analysis● Clean up cabling in injector area – some cables had
to go over the old MO rack. This can be improved during next shutdown
● Install sub-distribution boxes with diagnostics in Hall 3 Extension (shutdown time)
● Develop new concept of the 1.3 GHz generation to minimize phase drifts (if really needed)
07.10.2008 Krzysztof Czuba, Henning Weddig FLASH Seminar 28
Thank You For Attention!