First High Power Test of the ESS Double Spoke Cavity package Han Li On behalf of FREIA team FREIA Laboratory, Uppsala University 27th Oct. 2017

Responsibility for ESS Accelerator FREIA has developed the prototype spoke RF source and will validate the prototype spoke cavity and cryomodule with nominal RF power. Validation of high power RF amplifier (HPA) Validation of dressed spoke cavity Validation of cavity package fully dressed prototype cavity installed with FPC and CTS (in test cryostat) Requires the availability of one high power RF amplifier tested during phase 1 Validation of spoke prototype cryomodule complete prototype module (2 spoke cavities) Requires the availability of two high power RF amplifier test of switch-on procedures & LLRF Acceptance testing of series spoke cryomodules Han Li, 9th Jun. 2017

Recent RF Tests Test of cavity package. This test has the following goals: verify cooling procedures, verify power coupler conditioning procedure, coupler ability and performance, verify cavity intrinsic ability, accelerating performance, mechanical behaviour, verify LLRF ability and performance, verify the high power RF amplifier ability and performance in combination with the cavity and LLRF, verify cold tuning system (CTS) ability and performance, achieve nominal RF pulse. Typical measurements: RF behaviour during cool down, Coupler conditioning and cavity package conditioning, Achieve nominal gradient and nominal Q0, Cryogenic heat loads, Loaded Q-factor, eigen and external Q, Q0 = f(E) curve, Dynamic Lorentz detuning and mechanical modes, Field emission onset and multipacting barriers, Sensitivity to helium pressure fluctuations, Tuning sensitiviy.

The Self-excited Loop Test Stand (I) FREIA developed a test stand based on SEL for superconducting cavities under a pulse mode test at high power level. Help with the determination of cavity performance without tuner feedback system. FREIA SEL block diagram

The Self-excited Loop Test Stand (II) Combination with a DB RF station which can output up to 400 kW peak power. Developed digital phase shifter and gain-controller. Introduce interlock system for safety consideration. Introduce RF switch in order to manage a pulse operation mode. Developed SEL control and data acquisition system in LabView. Interlock and RF switch RF station SEL loop installed into a cabinet FPGA FREIA Labview SEL control system

Coupler conditioning (I) The warm and first cold RF processing was done using IPN Orsay’s system, Followed by the new FREIA conditioning system to verify its performance, All processes used a traditional signal generator driven loop, The warm RF processing procedure took about 30 hours, while the cold processing took roughly 14 hours, During warm conditioning, lots of outgassing occurred through the forward power region of 40-50 kW at short pulses, 120 kW forward power was reached with 2.86 ms pulse duration at 14 Hz.

Coupler conditioning (III) An automatic conditioning system, which consists of an acquisition system, a control system based on LabView software and feedback was developed at FREIA. Key parameters are primarily set. The main devices for the RF conditioning process are: Signal Generator Power Meter Vacuum Gauge Arc Detector Electron Detector Fast RF Interlock Switch Vacuum Pumping Cart Parameter value Loop control time (s) 1 Pulse repeat rate (Hz) 14 Vacuum upper limit (mbar) 1e-6 Vacuum lower limit (mbar) 5e-7 Initial pulse length (µs) 20 pulse length step 20 µs, 50 µs, 100µs, 200 µs, 500 µs, 1 ms, 1.5 ms, 2 ms, 2.5 ms, 2.86 ms

Coupler conditioning (IV) The FREIA automatic RF conditioning mainly collaborates with EPICS system. A new conditioning program based on EPICS is undergoing. The whole program consists of several modules, to make debugging easier and future upgrading more flexible. Initialize state Increase state Hold state Decrease state Vacuum upper limit state Fault state Ramp after fault state End state

Cool down Cool down of the LN2 shield from room temperature to 120 K took 21.5 hrs. The cavity package in HNOSS cooled down from 205 K to 4 K within an hour, From 4K to 2K it took an extra 20 minutes, Cavity frequency was checked during cooldown. 4.10 K/min 4.48 K/min 3.25 K/min Temp. Resonance frequency Frequency shift [MHz]  Frequency shift [%] 300 K 351.8939± 0.0001MHz -- 4 K 352.4004 ± 0.0001MHz -0.5065 ± 0.0002 0.1439 2 K 352.3725 ± 0.0001MHz -0.4786 ± 0.0002 0.136

Cavity Package conditioning Flattop gradient (MV/m) Cavity package conditioning was done by FREIA pulse SEL, 2.86 ms pulse with 14Hz repetition rate was used, Cavity package conditioning took about 30 hours, Three major multipacting regions have been found MP barrier Flattop gradient (MV/m) RF input power (kW) 1 4.5 - 4.8 22 - 30 2 5.2 – 5.7 35 - 48 3 7 – 7.5 67 - 76

Double Spoke Cavity Performance (Romea, 2017) Test of cavity package. ESS goal : 1.5×109 Low-field Q0 factor of 1.4×109@ 2 K Q factor of 2.8×108@ Eacc_peak=9MV/m Several multipacting regions were found Han Li, 9th Jun. 2017

Dynamic heat load (I) (Romea, 2017) Two different methods of dynamic heat load measurements to cross check the cavity performance: liquid helium evaporation (measured via the flowmeter placed after the sub-atmospheric pumps) the pressure rise method The cavity package dynamic dissipated power at 9 MV/m is about 12 W with 4% duty cycle, which is far higher than the expected value of 2.5 W. High radiation implies that MP or field emission happened in the cavity package, and it leads to high heat load and low Q factor. Eacc (MV/m) dynamic RF load (W) Test run Test method 9 10.71 1st run 2017-4-13 Flowmeter 8.98 13.16 2nd run 2017-4-25 13.35 Pressure rise 9.1 11.74 3rd run 2017-4-26 Han Li, 9th Jun. 2017

Dynamic heat load (II) (Romea, 2017) Considerations are focusing on the following five hypotheses: a contaminated FPC, particles generated during conditioning cryopumped on the cavity’s surface, FPC is not fully conditioned, a bigger impact on heat load than expected from the FPC into the cavity a combination of all these . Han Li, 9th Jun. 2017

Cavity Voltage Rising Time (I) Measured with SEL and pure square pulse @ 110 kW. Ramping up from noise to a flattop gradient of 9 MV/m, the cavity voltage takes about 800 µs with zero detuning. QL of 2×105, 148 kW@300 µs filling time Cavity voltage amplitude 14

Cavity Voltage Rising Time (II) Required filling power with different detuning and 300 µs filling time is studied. QL of 2×105, 1% more filling power @200 Hz detuning, 30% more filling power @ 1000 Hz detuning the simulation results for filling power as a function of detuning Han Li, 9th Jun. 2017

Cavity Voltage Rising Time (III) The system detuning usually varies during the filling stage in the real situation. By regulating a forward power from 152 kW to 350 W, the practical filling time is therefore in a range from 135 to 300 µs. Han Li, 9th Jun. 2017

Dynamic Lorentz Force detuning (I) Monitoring and manipulating the complex signal from cavity during the pulse, dynamic Lorentz force detuning at different gradient were studied. state space equation

Dynamic Lorentz Force detuning (II) Developed an FPGA-based LabView program for dynamic Lorentz force detuning. Frequency shift at peak accelerating gradient 9MV/m@ 2.86 ms pulse length is about 400Hz. A Loaded Q value of 1.8×105 from state space equation caculation is consist with the VNA measurement. ∆𝑓=400𝐻𝑧 𝑄𝐿=1.8×105

Mechanical Modes Stimulate the cavity by amplitude modulation . By sweeping the modulation frequency up to 800 Hz, the fit of mechanical modes was studied. Slow tuner is in fixed position . Simulation from IPNO N° Frequency 1 & 2 212 Hz 3 & 4 265 Hz & 275 Hz 5 & 6 285 Hz 7 313 Hz 8 to 11 315 Hz to 365 Hz 12 396 Hz Han Li, 9th Jun. 2017

Frequency Sensitivity to Pressure By closing both the inlet and outlet of the cryostat, checking the cavity frequency shift as a function of helium pressure from 20 to 40 mbar. pressure sensitivity = +27.1 Hz/mbar This result is consistent with the 28 kHz frequency shift measured during cool down from 4.2 K (~1030mbar) to 2 K (~20mbar).

Tuner Sensitivity Measurements Slow tuner is controlled by Lund system. Tuning sensitivity: 150 kHz/mm@2K

Test plan & Conclusions Maximum power of 120 kW and 9 MV/m accelerating gradient were reached, The quality factor of the Romea cavity package was determined 1.4×109 at low field and 2.8×108 at 9 MV/m, A high radiation of 6 mSv/h and a high heat load of about 12 W was observed, A practical filling time in a range from 135 to 300 µs can be obtained by using steps pulse profile, The resonant frequencies of Romea shift +500 kHz from 300 K to 4 K, - 27 kHz from 4 K to 2 K, The dynamic Lorentz detuning of about 400 Hz @ 9MV/m, Two mechanical modes of the cavity package, found at 343 Hz and 265 Hz, show a good agreement with simulations. High power test of high Beta elliptical cavity Test of spoke cavity cavities Romea and Giulietta will be installed in a prototype cryomodule with new fully conditioned FPCs at IPN Orsay, acceptance testing of the spoke prototype crymodule, Series test of 13 spoke crymodules at FREIA before delivering to ESS.