1. PSB C04 RF system Consolidation or upgrade? M. Paoluzzi – CERN BE-RF 11/23/20151

2. C02 : 4 cavities in 2 straight sections freq= 0.6-1.8 MHz Vgap=8 kV Rs≈350Ω seen by the beam C04 : 4 cavities in 1 straight section freq 1.2 3.6 MHz Vgap=8 kV Rs≈350Ω seen by the beam C16 : 4 cavities in 1 straight section freq 6.0- 16 MHz Vgap=6 kV Rs≈2 kΩ seen by the beam 11/23/20152

3. Consolidation foreseen C02, C16 Replacement of ITL, power supplies, cables (rebuilt the systems keeping the cavities and amplifiers) Possibly move AVC and Tuning loops to Low Level Consolidation foreseen C04 Same as above but power stage replacement (improve mean available power and current for beam loading compensation) Is it the time to envisage a technology change? 11/23/20153

4. Narrow band Tuned systems Highly optimized Almost no room for Vgap increase Further reduction of Rs very difficult 11/23/20154

5. Finemet LEIR Cavity A single cavity covers the range 0.35÷5 MHz Cavity length 0.5 m Gap voltage 4 kV (tested at 5 kV) No tuning AVC integrated in Low Level But… … diameter incompatible with PSB overlapped rings 11/23/20155

6. Finemet in PSB Maximum outer diameter OD=330mm Inner diameter ID=200mm Standard Finemet thickness 25mm Two kind of material: Standard LEIR like : FT3M (data available from small rings) New low loss : FT3L (Thanks to KEK and Chihiro Ohmori for providing data) 11/23/20156

7. Air cooling test on small rings (OD=133mm) : - Simple blower on each ring side - With 500kW/m 3 t MAX ≈70°C - Maximum inside temperature unknown - Limit temperature - Suggested from Hitachi t MAX ≤100°C - KEK booster experience t MAX ≈100°C - J-PARC experience t MAX ≈200°C for few hours 11/23/20157

8. Minimum peak voltage 380V for FT3L 300V for FT3M @ 1 MHz Note: f Injection ≈1 MHz with LINAC4 One C04 cavity contains 50 rings In the range 1÷4 MHz : V GAP =15÷17 kV with FT3M V GAP =19÷29 kV with FT3L 11/23/20158

9. Basically working on a capacitor Almost no difference with increased number of rings 11/23/20159

10. Alternative approach based: use two rings (FT3L) and 1 gap Each ring can accept 500pF and still cover the 1÷4 MHz band. R Pmin is ≈100 Ω (per ring) P≈1500W for 700V across two rings This can be supplied by a solid-state amplifier (in PSB solid state amps used since 1987!!!) 11/23/201510

11. Beam current 11/23/2015 w/o acceleration, the h1 bunch length: φ L  138 o (2.4 rad or 220 ns h1 @ej.) Total charges per bunch= 2 E13 h1 (sinus shape approx.) For a pure h1 or h2 beam, the beam peak current is 22.7 A @ ej 1 st rf harmonic : Î H1 = 9.56 A 2 nd rf harmonic: Î H2 = 6.06 A φ L, T L Sine shape approximation 11 Slide provided by A. Blas CERN BE-RF

12. MRF157 Mosfet 600W up to 80MHz To provide 800W (2 units push-pull) I D-peak ≈32A I D-DC ≈10A 11/23/201512 MRF151 Mosfet 150W to 175MHz To provide 200W (2 units push-pull) I D-peak ≈12A I D-DC ≈4A

13. Shunt resistance of each gap Rs≈200 Ω To get the present 8kV need 12 gaps Total impedance seen by the beam Rs tot ≈2400 Ω Impedance of present cavities (including FB) ≈350 Ω We must reduce Rs tot and compensate beam loading with: ≈20 dB RF feedback seems possible with 10 ns delay in the loop. 1 turn feedback: difficult (nevertheless possible) because revolution time is only ≈500 ns at extraction. Combination of both 11/23/201513

14. The cavity could be used as combiner for push-pull operation and driven by 2 amplifiers similar to the LEIR 1kW drivers. The total power could easily be brought to 2 x 2 kW simply increasing the DC supply by 15%. The 120W (200W) power module delay is 12 ns and the 9 ways combiner 4ns. Both could be reduced and 10 ns total seems within reach. 11/23/201514

15. 11/23/201515 Simulated system response: 1 Finemet ring driven by 9 combined 200W modules and feedback including delay. Amplification chain response. 200W module. 1 ring impedance.

16. Advantages Modular system. Reduced machine space and/or increased installed voltage. Possibility of installing spare cavities as hot back-up. No tuning loop. Simplified solid-state system. Risks Additional new system at restart with LINAC4. New technology and configuration. Complete new design. 11/23/201516

17. How could we proceed… 2010-2011 Design and produce a first section. … if successful then: 2012 produce and install and test a full set of cavities for the required 8 kV. They should fit in the place of one of the present C04 cavities. 2013 Tests with beam (1 ring and C02 system still in place) … if successful then: 2014-2015 Proceed with full scale construction and C02/C16 renovation 2016 Installation … but for security 2011 Design a first C04 improved RF power amplifier. 2012 Produce and install C04 improved RF power amplifier. 2013 Tests with beam (1 ring) 2014-2015 Depending on 2013 tests start C02/C04/C16 renovation or stop C04 renovation it and proceed with new system. 2016 Installation 11/23/201517