RF considerations for H. Fitze in behalf of the PSI RF group Workshop on Compact X-Ray Free Electron Lasers 19.-21. July 2010, Shanghai Agenda Introduction RF-Gun Development C-band development Summary 2/ 25
Baseline Layout Injector S-Band 2998.8 MHz (21xf b ) X-Band 11995.2 MHz (84xf b ) Main LINAC C-Band 5712 MHz (40xf b ) f b =142.8 MHz 3/ 25 RF System Tolerance ballpark: amplitude: 10-5 -10-4 phase: 0.03 250 MeV Injector RF Systems CTF gun INFN 1 cell deflector 1 2Π/3 TW 4 m long INFN 5 cell deflector 2 TW1 TW2 TW3 TW4 25 kw babyklystron from CERN 3 S-band 35 MW 4.5 μs Klystron TH2100E 3 3 3 3 12 S-band Mode 1: 45 MW 4.5 μs Mode 3: 60 MW 1.2 μs Klystron TH2100L X-band 50 MW 1.5 μs XL5 SLAC 3 S-band 7.5 MW 4.5 μs TH 2157 4/ 25
5/ 25 project Time Line R&D Schedule Phase 1 & 2 R&D and prototyping of large number components (RF systems, undulators,...) R&D and prototyping of special components (special instrumentation, injector, feedback,...) R&D and prototyping phase 2 Accelerator Buildings Planning and building permits Component procurement accelerator and ARAMIS FEL Buildings and tunnel construction Technical infrastructure Installation accelerator and ARAMIS FEL Commissioning accelerator and ARAMIS FEL Procurement ATHOS FEL Installation ATHOS Commissioning ATHOS Year 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Milestones Science Case Request to Report ETH council Conceptual Design Report Request to parlament Technical Design Report Begin civil construction Buildings and Infrastructure ready First beam phase 1 Start user operation First beam phase 2 6/ 25
RF R/D Projects for Status Due Remarks S-band RF Gun Reviewing mechanical drawings 2011 LCLS, PHIN X-band acc. structure Machining cups 2010 Collaboration with CERN C-band acc. structure Brazing tests, finalizing RF design 2013 C-band pulse compressor Mechanical design started 2012 Collaboration with INFN C-band deflecting cavity Not yet started 2015 C-band modulator Procurement in progress 2011 SCN C-band klystron Procurement in progress 2011 Toshiba LLRF S-band prototype ready 2011 PSI development 7/ 25 Pulsed DC Gun Charge Energy ε x 180 pc 5.7 MeV 0.5 μm 8/ 25
Pulsed DC Gun 9/ 25 10 / 25
RF Gun RF Gun Beam Laser 1.5 1 EFY undef norm EEY def norm 0.5 0 0 0.05 0.1 0.15 0.2 0.25-0.5-1 -1.5 11 / 25 RF Gun (cont.) t pulse (μs) 5 4 3 5 3.75 2 1 2.5 6.25 7.5 5 1.25 Dissipated power in kw (400 Hz) 7.5 6.25 3.75 8.75 2.5 11.25 10 12.5 5 15 13.75 10 8.75 12.5 11.25 6.25 3.75 17.5 16.25 7.5 18.75 2.5 22.5 21.25 20 15 13.75 10 8.75 5 23.75 27.5 26.25 25 22.5 18.75 17.5 16.25 12.5 11.25 7.5 6.25 3.75 31.25 28.75 30 21.25 20 15 13.75 23.75 10 8.75 5 30 27.5 26.25 25 41.25 38.75 40 37.5 33.75 35 32.5 21.25 18.75 17.5 20 16.25 28.75 15 12.5 11.25 6.25 36.25 50 75 100 125 150 E acc (MV/m) 7.5 E acc (MV/m) 140 120 100 80 60 40 20 LLRF: amplitude program 0 0 1 2 3 4 5 6 Time (μs) 12 / 25
C-band LINAC Module Modulator LLRF Main LINAC LINAC modules Modulator Klystron Pulse compressor Accelerating structures Waveguide splitter Waveguide loads # 26 26 26 26 104 78 104 Pulse- Compressor 40 MW, 2.5 µs 116 MW 26.5 MV/m 26.5 MV/m 26.5 MV/m 26.5 MV/m 13 / 25 C-Band LINAC structure development RF design Const. gradient structure with J-type coupler (5712 MHz, 2π/3). Detuning of the structure might be necessary (two bunch operation) Mechanical / fabrication concept No tuning of the structures ( precision manufacturing). Conventional joining technology (e.g. brazing instead of diffusion bonding) Develop technology in house and together with external company. ~ 2m 14 / 25
Brazing Development Steps 1. Brazing of two regular cups evaluate brazing parameters check mechanical precision 2. Brazing of two UP cups check RF precision 3. Brazing of a stack of UP cups (max. length 800 mm) check RF properties check cup alignment RF high power tests (using mode launchers) 4. Brazing of full structure (cannot be done at PSI) check everything 15 / 25 16 / 25
Ansoft LLC 180.00 135.00 90.00 45.00 0.00-45.00-90.00-135.00 HFSSDesign1 Curve Info ang_deg(s(1,2)) Setup2 : Sw eep1 ANSOFT RF Pulse Compressor Electric Field S-Band Barrel Open Cavity (BOC), CERN S-Band BOC, CERN XY Plot 1 ang_deg(s(1,2)) [deg] Magnetic Field -180.00 5.660 5.680 5.700 5.720 5.740 5.760 SPring-8 Freq [GHz] 17 / 25 Energy gain vs. bunch arrival time energy gain per AC (ev) disk loaded SKIP 60000000 disk loaded BOC 59000000 double_rounding SKIP 58000000 double_rounding BOC 57000000 cup SKIP 56000000 cup BOC 55000000 54000000 53000000 52000000 51000000 50000000 49000000 48000000 47000000 46000000 45000000 2300 2350 2400 2450 2500 2550 2600 t bunch (ns) P RF_Klystron = 41 MW 18 / 25
Klystron: Toshiba E37202 Frequency Beam Voltage Cath. Current Efficiency RF drive RF output pk. RF output av. Beam perv. Pulse rep. rate Used in SPring-8 5712 ±5 MHz 370 kv 344 A 43 % 500 W 50 MW 12.5 kw 1.5 μav 3/2 60 Hz Will buy E37202 to be able to start high power testing in the C-band RF test stand. Klystron will arrive in spring 2011. Need upgraded klystron for a pulse repetition rate of 100 Hz. 100 Hz klystron should be ready in early 2012. 19 / 25 Modulator: ScandiNova First operational experience in the 250 MeV injector indicates usefulness of ScandiNova solid state modulators. Will buy ScandiNova modulator for C-band RF test stand. Should be operational in spring 2011. 20 / 25
LLRF System 21 / 25 RF gun (FINSS) hw: clone of DESY type digital LLRF system (SIMCON DSP) no pulse to pulse feedbacks so far try to understand the measured data first! new digital LLRF system for the 250 MeV injector: hardware common digital hw, VME controls interface; frequency specific front ends (3 GHz, 12 GHz) systems (S band) ready for commissioning digital hw: FPGA (Virtex5) processing board with 16 bit/125 MS/s ADC, DAC 12 GHz RFFE development: design phase started software controls integration completed 10 Hz rep. rate for beam operation 100 Hz rep. rate for modulator tests (data transfer to ctrl. system reduced to 10 Hz) pulse to pulse feedbacks in preparation performance: σ ampl. ~ 5 10-4 (intra-pulse) σ phase ~ 0.03 ctrl. interface (VME system) temporary RF generator dig. processing hardware 6 channel S band receiver LO generation / vector modulator 22 / 25 performance in lab (S band) (detection bandwidth: 7 MHz) performance: σ ampl. ~ 1.2 10-4 (intra-pulse) σ phase ~ 0.014
RF gun: RF pulse-to-pulse stability First measurements Phase Forward Power [ ] 19-March-2010 Gun Temp. [ C] Klystron gun (FINSS) Pulse-to-pulse RF phase of the incident power during 10 minutes (10 Hz rep. rate) Phase of every pulse averaged over 400 ns (40 samplings) no RF feedback so far! In blue: filtered data for future RF feed back 2 min. period phase oscillation: correlated with water temperature oscillations slow phase drift: correlated with LLRF electronics ambient temp. Reference Phase [ ], 3 GHz LLRF Ambient Temp. [ C] future strategy: better ambient temperature stability and/or temp. stabilized LLRF electronics necessary reference tracking beam based feedbacks (later stage) 23 / 25 RF gun: RF pulse-to-pulse stability Phase stability [ ] (forward power) Max. pulse to pulse phase jitter: 0.02 (rms) 29-March-2010 Amplitude stability [relative] (forward power) phase change (deg RF) 0.02 klystron high voltage stability 3.6 10-5 Max. pulse to pulse amplitude jitter: 0.019% (rms) rel. amplitude change klystron high voltage stability Pulse-to-pulse data: standard deviation over the last 40 pulses (rep. rate 10 Hz). At each pulse the data is averaged over 400 ns. 24 / 25 29-March-2010 klystron high voltage stability: < 3.6 10-5 (rms) pulse-to-pulse ampl. jitter not yet understood (pre-amplifier?) klystron operated in non-saturation
Summary RF group is currently installing and commissioning the S-band linac in the injector test facility. Development, engineering and fabrication of an S-band RF gun is in progress. Design, engineering and fabrication of a C-band linac structure and a BOC RF pulse compressor is in progress. Setup of a dedicated C-band test stand has been started (should be operational mid 2011). It is foreseen to test a complete C-band linac module. First PSI LLRF system (S-band) is operational. C- and X-band are still missing. Development of an optical timining/synchronisation system is in progress (not covered in this talk). 25 / 25