CW Measurements of TESLA and Gun Cavities with the Cornell LLRF System at HoBiCaT Axel Neumann SRF Science and Technology Helmholtz-Zentrum Berlin for a collaboration by Cornell University and HZB: S. Belomestnykh*, J. Dobbins, R. Kaplan, M. Liepe, C. Strohman W. Anders, R. Goergen, J. Knobloch, O. Kugeler (*now BNL)
Lunar Landing Research Facility, Hampton, Virginia, Another perspective of LLRF: NASA Lunar Landing Research Facility, Hampton, Virginia, NASA Langley Research Center
Motivation: Field control for SRF cavities in Energy Recovery Linacs Visit: http://erl.chess.cornell.edu/ HZB ERL: BERLinPro SRF Photo-injector Main linac cavity (with JLAB style HOM dampers) High beam current, low beam power Riemann et al. IPAC 2011 SRF booster cavity (Cornell type) High beam power and high beam current Parameters Beam energy 50 MeV Max. current 100 mA Nominal bunch charge 77pc Max. rep. rate 1.3GHz Normalized emittance < 1mm mrad
Overview and challenges for LLRF: Low beam-loading (zero?) allows high QL operation CW operation: Microphonics, peak events? High cw gradient: 20 MV/m Ponderomotive instabilities High cryogenic dynamic losses, helium bath stability Beam transients during ramping to 100 mA, how to handle? Low current diagnostic mode…. Residual beam-loading due to beam losses non- perfect recovery, time jitter? Combine microphonics compensation with LLRF at high QL for a multi-cell cavity TESLA cavity as testbed for cw driven ERL main linac cavity II. Synchronization to cathode laser CW operation at strong microphonics Strong response to LHe pressure variations First version without tuning scheme Field ramp at strong Lorentz-force detuning Frequency deviates from design frequency 1.6-cell SC gun cavity as first step towards ERL photo- injector in collaboration with: JLAB, DESY, A.Soltan Institute, MBI, BNL for beam dynamic studies
Collaboration HZB with Cornell University at HoBiCaT Goals of the measurements Determine the ultimatively achievable field stability: What are the maximum loop gain settings? What is the optimum cavity bandwidth for ERL main linacs (with respect to stability and forward power)? Determine microphonics detuning level, peak events and spectrum How to ramp cavity field at small bandwidth in the presence of Lorentz force detuning Combine detuning and LLRF field control for a most reliable and stable system Detailed description of the LLRF hardware and performance at Cornell see: Cornell Lab talk by Vivian Ka Mun Ho on Monday
Testbed: The Cornell LLRF system at HoBiCaT Reference oscillator: 1.3 GHz, below 36 fs integral time jitter of 10 Hz – 5 MHz sideband spectrum Gun cavity TESLA cavity TESLA cavity, courtesy DESY 17 kW CPI IOT 6xADCs (16 bit): Pfor,Pref,Ptrans, Beam signal,… Slow DAC/ADC interface 2xDACs 2nd generation system Saclay I tuner with high voltage piezos Here piezo in PI loop only
Power requirements and parameter space Studied within this work Pros: High QL: low forward power for a given field level, reduction of thermal stress for RF transmission line But: Effective detuning is a convolution of the detuning spectrum with the cavity response (bandwidth) Cavity transfer function itself altered by controller settings (feedback gains) QL 7-cell cavity, 20 MV/m, Ib=0
Detuning spectrum versus bandwidth (TESLA cavity) For two different tuning schemes (Saclay I and INFN Blade) open loop measurements of microphonics vs. QL were performed Both tuners showed to have different transfer functions and thus detuning spectra QL,Saclay: 3.107-4.108 QL,Blade: 7.105-2.107 Blade: Mechanical eigenmode at 300 Hz, vacuum pump freq. Saclay: Excitation of 1st mechanical eigenmode sets in
Detuning and transfer functions Measured detuning spectrum Mechanical resonances Turbo pump lowest mechanical eigenmodes at 21 and 36 Hz group delay between 240-300 ms piezo tuning range above 1 kHz
Time domain results: Proportional gain scans Residual error evolves with 1/(1+KP) Upper limit by noise of RF reference system At low gain strong coupling of detuning and amplitude via Lorentz force detuning Lorentz force effect
Parameter studies: Optimum gain and cavity bandwidth log(sf) Best results: 5.107 0.008° 1.108 0.0093° 2.108 0.0236° QL=5.107, f1/2=13 Hz QL=2.108 , f1/2=3.25 Hz 9 cell TESLA cavity Eacc= 10 MV/m Tbath= 1.8 K PI piezo loop 8/9-p filter optimized QL=1.108, f1/2=6.5 Hz LF detuning IOT beam instable Areas with sf>0.1 were blanked out
Limitations for combining LLRF and piezo tuner control? Amplitude and phase spectrum vs. proportional gain KP Detuning information Amplified phase noise? sf=6 Hz sf=9.3 Hz sf=9.6 Hz Spectrum of calculated tuning angle by Pt and Pf Simulation Noise dominated
Best results achieved / Summary for main linac QL sf (Hz) sf (deg) sA/A Pf (kW) 5.107 9.5 0.008 1.10-4 1.106 1.108 7.9 0.009 2.10-4 0.595 2.108 4.2 0.024 3.10-4 0.324 But cavity was limited by field emission to 10 MV/m. We plan to repeat test at 20 MV/m with a good and clean cavity. Piezo loop only operated at low frequency PI mode, still plan to implement adaptive feed forward (see DOI10.1103/PhysRevSTAB.13.082001) The low noise reference source broke down when tests were started, performed test with a R&S standard frequency synthesizer results limited by noise Klystron ripple loop was not implemented, the IOT’s power supply has intrinsically low noise, but very non-linear behavior of IOT at power levels below 1 kW Next: Gun cavity
Design: J. Sekutowicz* TTF-III FPC: QL= 1.109-6.106 Cavity fabrication at Thomas Jefferson Lab (P. Kneisel, Proc. PAC 2011) Fundamental power coupler port Stiffening ring Status Anfang des Jahres: Large grain cavity backwall for cathode Cavity half- cells Pick-up Port Passive stiffening System: “Spider“ Design: J. Sekutowicz* „Helium vessel endplate“ TTF-III FPC: QL= 1.109-6.106 including 3-stub tuner Cavity in helium vessel Pb cathode on back plane *Sekutowicz et al., Proc. PAC 2009
Mechanical design: Countermeasures to increase field stability Beam quality dominated by field stability Field stability in SC cavity: Avoid detuning (deformation) of the cavity Combined FEM mechanical and Electro-magnetic field simulations low deformation (detuning) design Backwall stiffening Detuning in Hz/mbar F Lorentz-force detuning: Df/Epeak² = 1.33 Hz/(MV/m)² 114 Neumann et al., Proc. Linac 2010
Synchronizing Laser and LLRF without tuner Both, GDR LLRF and Laser PLL synchronized to reference follow slow frequency drifts of gun cavity. Tested laser PLLs bandwidth by lock-in measurement F Laser+PLL 1/N Loop filter Gain Use modified piezo loop to add slow PLL loop for drift stabilization of cavity and to ramp field First beam 21st of April 2011 on YAG viewscreen
QL Epeak sf sF sA/A Dfpeak (Hz) 6.6.106 20 MV/m 7.0 Hz 0.017 deg Field stability Microphonics spectrum Mechanical eigenmode QL Epeak sf sF sA/A Dfpeak (Hz) 6.6.106 20 MV/m 7.0 Hz 0.017 deg 1.2.10-4 25.6 Hz 1.4.107 12 MV/m 5.0 Hz 0.02 deg 1.5.10-4 20 Hz Using Cornell LLRF system+ slow PLL loop
Thank you for your attention Summary for gun cavity Adapted Cornell’s LLRF system within a few days to be operated with gun cavity without tuner System routinely ramps cavity to peak fields on cathode of 22 MV/m and even more Piezo tuning algorithm used in lowpass PI configuration to follow cavity drift by reference Laser PLL showed no sign of loosing lock so far Within the framework of the performed beam measurements the set up is stable Jacek Sekutowicz (DESY) and Peter Kneisel (JLab) are working on a new cavity with tuning scheme adopting the Saclay I tuner for the 1.6 cell design Thank you for your attention
Backup transparencies
Possible reasons for instability IOT response time and stability: Example: QL=2.108, 1.5 Hz detuning Eacc drops to 9.07 MV/m: (10 MV/m)²-(9.07 MV/m)² Lorentz force detuning: ~17.6 Hz LLRF needs to compensate by 30 times RF power At high loop gains the requested power even increases on short time scales Beam of IOT becomes unstable leading to an RF trip The cure: Combine detuning control and LLRF to minimize „high power“ events Operation at high QL possible, thus saves RF power installations!
Extension of the HoBiCaT Cavity Test Facility Laser hut Laser beamline C o n t r l ro m HoBiCaT cryostat Diagnostic e- beamline RF amplifier
Gun with Diagnostic beamline at HZB SC Solenoid Stripline BPM, ICT THz diagnostics SQ Faraday Cups SC Cavity Dipole magnet Viewscreens HoBiCaT cryostat Mass Spectrometer First time operated fully SC photoinjector ensemble (SC Cathode, Cavity, Solenoid) Source/upgrade for CW low current machines (POLFEL, XFEL, FLASH)
EM design: Highest fields at cathode region SC RF Gun0 Pick-up probe Excited p-TM010 mode Original design by Jacek Sekutowicz* Lead cathode, Tc= 7.2 K Input coupler EM design: Highest fields at cathode region SC lead cathode on half- cell backwall: QEPb~10.QENb Study beam dynamics at short pulses, ERL parameter range Frequency p-mode 1300 MHz Epeak/Eacc 1.86 Hpeak/Eacc 4.4 mT/(MV/m) Geometry factor 212.2 W R/Q (linac, b=1) 190 W *Sekutowicz et al., Proc. PAC 2009
Ibeam=50 nA at Ecath=20 MV/m Kamps et al., Proc. IPAC 2011 Neumann et al., Proc. IPAC 2011 Beam energy First beam 21st of April 2011 on YAG viewscreen Beam current (nA) Laser phase (deg) 3-4 ps pulse length during extraction Bunch charge: 5-6 pC Ibeam=50 nA at Ecath=20 MV/m at 8 kHz Laser rep. Rate, l=258 nm best QE=1.10-4
Reconstruction of the transfer function Fit by par. systems of 2nd order Include response of higher modes at lower frequencies 20 modes needed for the range shown Complex system complicates model-based design (e.g. Kalman filter) Transfer function implemented as table Tuning relevant range