commissioning of the high current erl injector at cornellcommissioning of the high current erl...
TRANSCRIPT
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
Commissioning of the High Current ERL Injector at Cornell
Florian Loehlfor the Cornell ERL team
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
photocathodeDC gun
buncher
cryomodule
beam dump
experimental beam lines
deflector
Design parameter:Nominal bunch charge 77 pCBunch repetition rate 1.3 GHzBeam power up to 550 kWNominal gun voltage 500 kVSC linac beam energy gain 5 to 15 MeVBeam current 100 mA at 5 MeV
33 mA at 15 MeVBunch length 0.6 mm (rms)Transverse emittance < 1 mm-mrad
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
Initial thermal emittance measurements
Measured normalized beam emittance at ~fC bunch charge (5 MeV):
0.2 to 0.4 mm mrad (in both planes)
Good agreement with predictions from thermal limit at cathode and utilized laser spot size.
Next step: optimizing injector for 77 pC operation• Implemented fast (~5 s), “single button” measurement• Will be used for parametric optimization of the injector
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
Comparison of beam measurements with simulations
Fixed slits phase space measurements• Corrector coils for beam scanning• 10 to 20 micron precision slits• 1 kW beam power handling
Good agreement with theory givesconfidence that the very small simulated emittances can be achieved.
From
Phy
s. R
ev. S
T-A
B 1
1, 1
0070
3 (2
008)
Beam properties at the cathode
Phase space measured after the beam was transported through theaccelerator.
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
Beam Orbit Control
Implemented global beam position feedback which uses all BPMs, corrector, and dipole magnets.
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
Beam Orbit Control
Implemented global beam position feedback which uses all BPMs, corrector, and dipole magnets.
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
Transverse Deflecting Cavity
beam energy
time
streakedbeam
unstreakedbeam
Number of cavities 1 Max transverse kick voltage 200 kV Max RF power 3.8 kW Average power 200 W Pulse duration 60 µs Max rep. rate 1 kHz
Pumping port
Protrusion
Tuner and tuner mechanism
Water cooling channel
Input coupler
Beam pipe
Field probe
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
Initial high current run
0
2
4
6
8
10
12
10:00:00 PM 10:30:00 PM 11:00:00 PM
beam
cur
rent
(mA)
• Achieved maximum current is about 9 mA• Main limitations: - Gun high voltage instabilities
- Laser amplitude / position instabilitiesWork on solving the stability issues is in progress
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
Strange beam response of 250 kV beam
Steering the beam differently through the cryomodules changed the beam shape (no RF field in cavities)
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
Strange beam response
horizontal corrector (A)
verti
cal c
orre
ctor
(A)
verti
cal b
eam
pos
ition
(mm
)
horizontal beam position (mm)
cryomodule beam position measurement
correctorpair
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
Localizing stray fields in the cryomodule with DC coupler-kicks
horizontal position (mm)
verti
cal p
ositi
on (m
m)
electrical field vectors
+-
horizontal position
+ +
dipole-like field quadrupole-like fieldGeneration of electrical DC fields in the coupler regions of the SRF cavities
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
Localizing stray fields in the cryomodule with DC coupler-kicks
coupler 5coupler 4coupler 3coupler 2coupler 1
hor. corrector strength
hor. corrector strength
hor. corrector strength
hor. corrector strength
hor. corrector strength
vert.
cor
r. st
reng
th
deflection of the beam after the cryomodule due to the coupler field
cryomodule beam position measurement
correctorpair
dipole-like coupler kick
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
Localizing stray fields in the cryomodule with DC coupler-kicks
coupler 5coupler 4coupler 3coupler 2coupler 1
cryomodule beam position measurement
correctorpair
quadrupole-like coupler kick
deflection of the beam after the cryomodule due to the coupler field
hor. corrector strength
hor. corrector strength
hor. corrector strength
hor. corrector strength
hor. corrector strength
vert.
cor
r. st
reng
th
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
Localizing stray fields in the cryomodule with DC coupler-kicks
3000 3500 4000 4500 50000.00
0.05
0.10
0.15
0.20
Center of 3rd HOM absorber
Coupler 2
Coupler 3
Coupler 4
Determined the origin of the stray fields:
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
In-situ demagnetization
Warm up and in-situ demagnetization removed stray fields
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
Stray fields
horizontal corrector (A)
verti
cal c
orre
ctor
(A)
verti
cal b
eam
pos
ition
(mm
)
horizontal beam position (mm)
• Stray fields reappeared after a beam loss in the cryomodule• Coupler conditioning changed the stray fields Charging up of HOM absorbers!
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
HOM absorbers
17
Cooling Channel (GHe)
Flange to Cavity
Flange to Cavity
RF Absorbing Tiles Shielded
Bellow
Total # loads 3 @ 78mm + 3 @ 106mmPower per load 26 W (200 W max)HOM frequency range 1.4 – 100 GHzOperating temperature 80 KCoolant He GasRF absorbing tiles TT2, Co2Z, Ceralloy
Antennas to study HOM spectrum
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
Charging up of HOM absorbers
Low conductivity of HOM absorber tiles: Can hold charge for many days / weeks! Worst offender: Ceramic 137Zr10, followed by ferrite Co2Z and TT2
Small beam loss charged up absorber tiles -> kV electric fields at beam position!
Opposite side Cu coated
Uncoated between electrodes
Opposite side Cu coated
Uncoated between electrodes
1.E-111.E-101.E-091.E-081.E-071.E-061.E-051.E-041.E-031.E-021.E-011.E+001.E+011.E+02
70 170 270 370 470
σ[1
/Ω/m
]
T [deg K]
TT2 #2TT2 #1TT2 #3Co2Z #2Co2Z #3Co2Z #1137Zr101 sec time const
>1 second time const
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
Removing inner tiles
Stress relief slots
Beamside tiles removed
Consequence of low resistivities of absorber materials:• Completely removed ceramic 137Zr10• Tried gold coating of TTE absorbers but coating may fall off Removed all tiles from the inside of the HOM absorber
Found one loose tile during cryomodule disassembly• Thermal stress tests confirmed
this problem Solved by cutting stress relief
slots in the tiles
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
Power limitations with modified HOM absorbers
2000 2500 3000 3500 4000 4500 5000100
102
104
106 X: 2630Y: 2.75e+005
R/Q
*Q [Ω
]
f [MHz]
X: 3751Y: 364.8
X: 3904Y: 692.3
100 mA, 1 kW limit
Blue: inside and outside ferritesRed: outside ferrites only
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
Power limitations with modified HOM absorbers
2000 2500 3000 3500 4000 4500 500010-4
10-3
10-2
10-1
100
X: 2630Y: 0.06555
f [MHz]
X: 3751Y: 0.6855
X: 3904Y: 0.06082
Pw
all /
Pto
tal
Blue: inside and outside ferritesRed: outside ferrites only
Power loss in the metal walls
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
Cavity Quality Factors
109
1010
0 2 4 6 8 10 12Eacc [MV/m]
Q
August 2008April 2009
Averaged quality factor Q0 vs. Eacc for all cavities together Had difficulties with low cavity
quality factors
• Q factors degraded over time Q disease?
During the rebuild, all cavities were high pressure rinsed Q restored to 1.6 x 1010 at 1.8 K no Q disease cavities were possibly contaminated with particles?
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
Current status of the injector cryomodule
• Cryomodule is rebuilt and back in the injector
• Cooled down to 4 K • 2 K cool-down planned for next Monday
Ready to see beam in the next weeks!
FLS2010 Workshop, Stanford, March 1-5, 2010 Florian Loehl (Cornell University)
Conclusion and Outlook
Charging up of the HOM absorbers caused difficulties during the first commissioning Rebuild cryomodule during the last 6 months and
removed problematic absorbers
Still, many critical systems could be successfully commissioned and prepared.
• Measured thermal beam emittance as expected• Could increase beam current to 9 mA for short times
(limitations understood and being worked on)
First beam operation after cryomodule rebuilt expected end of March
• Emittance optimization at 77 pC• Work on high current beam operation