SPS Wideband Kicker Electromagnetic Design Manfred Wendt CERN

SPS Wideband Kicker – Electromagnetic Design – Manfred Wendt CERN With a lot of help and input from the LARP collaboration J. M. Cesaratto, J. D. Fox, C. H. Rivetta, SLAC D. Alesini, A. Drago, A. Gallo, F. Marcellini, M. Zobov, INFN-LNF S. De Santis, Z. Paret, A. Ratti, H. Qian, LBL H. Bartosik, W. Hölfe, C. Zannini, CERN

Slotted Waveguide Stripline Kicker • Periodic, quasi-TEM travelling-wave structure – Phase velocity vp close to c 0 – Combines stripline (TEM) and slotted-waveguide (TE) kicker characteristics Stripline electrode, 65. 5 x 5 mm, – Allows broadband operation, ~1000 mm long here: DC – ~1 GHz • Figures of merit load (50 Ω) beam – Bandwidth (BW), rise / fall time (tr, tf) – Transverse shunt impedance Zsh=f(ω) V± Beam pipe, 132 x 52. 3 mm load (50 Ω) WR-430 waveguide, 109. 22 x 54. 61 mm, ~1000 mm long 20 th September 2016 – Wideband Kicker – Electromagnetic Design (M. Wendt) 40 coupling slots, 70 x 12. 5 mm, 25 mm pitch Page 2 V∓

Design Options (LARP) • Evaluation of the parameter space – Various types, slotted WG only, stripline with and w/o WG ridge Ø Also compared to resonant kickers and stripline arrays – Slot dimensions – # of slots – Stripline width • More: see design report 20 th September 2016 – Wideband Kicker – Electromagnetic Design (M. Wendt) Page 3

Mechanical Dimensions thick sw sl bw helec pitch tinner wh bh 20 th September 2016 – Wideband Kicker – Electromagnetic Design (M. Wendt) Opti beam pipe bw 132 bh 52. 3 wave guide ww 150 109. 22 wh 50 54. 61 strip winner line tinner 80 65. 5 5. 0 helec na 25 sl 80 70 sw 12. 5 thick 1. 0 3. 0 # 40 40 pitch 25 25 slots ww w inner LARP Page 4

Eigenmode Analysis • Two TEM-line modes – Eignemode analysis with periodic boundary conditions Ø E-field phase advance ranges 2. 50 -450 • • – Σ and Δ mode Mode 2 (Δ) couples magnetically Phase-beam velocity slippage: – ~310 ps @ 1000 mm length, βsignal = 0. 915 (@ f = 1. 2 GHz) 20 th September 2016 – Wideband Kicker – Electromagnetic Design (M. Wendt) Page 5

Wakefield Analysis • Result differs from LARP – Zsh = ~30 kΩ vs. ~5 kΩ for similar dimensions!? beam axis (x=0 mm, y=0 mm) integration path (x=+1 mm, y=0 mm) 20 th September 2016 – Wideband Kicker – Electromagnetic Design (M. Wendt) Page 6

Particle-in-Cell (PIC) Analysis Port 2 signal V- Horizontal deflection, 1800 mm downstream, to a 50 ns long rectangular kicker pulse Port 1 signal V+ Leading edge of the deflected beam response trajectory DC-line beam 20 th September 2016 – Wideband Kicker – Electromagnetic Design (M. Wendt) Page 7

Details of the PIC Analysis • • • Quasi DC proton beam – Point source, E = 26 Ge. V, I = 0. 1 A, tr = 1 ns Kicker electrode stimulus signals – Differential rectangular pulse – tpulse = 50 ns, tr = tf = 100 ps, tdelay = 5 ns, ΔV = 5 MV Beam response, Δs = 1800 mm downstream – xpos = 15 mm -> θkick ≈ 8. 33 mrad -> Zsh ≈ 47 kΩ (!? ) – tr = tf ≈ 200 ps -> equivalent BW > 1. 5 GHz • Ø Simulation BW = 10 GHz Ø Some overshoot and ringing Ø Phase velocity < c 0 does not affect DC beam Ø Practically the bandwidth may be limited tslip ≈ 310 ps -> BW ≈ 1. 2 GHz This simulation was performed with a kicker structure of slightly different dimensions than the latest optimization – sl = 68 (70) mm, helec = 27. 5 (25) mm, thick = 1. 0 (3. 0) mm Ø Very time consuming computation! 20 th September 2016 – Wideband Kicker – Electromagnetic Design (M. Wendt) Page 8

PIC Analysis: Hy-Field Flatness vertical horizontal H-field at 40 ns longitudinal 20 th September 2016 – Wideband Kicker – Electromagnetic Design (M. Wendt) Page 9

Impedance Analysis Mechanical model (STEP file) EM model with parametric dimensions and dependencies!! 20 th September 2016 – Wideband Kicker – Electromagnetic Design (M. Wendt) Page 10

Impedance Matching volume of uncontrolled transmission-line impedance (red boundaries) coaxial feedthrough sections with well controlled impedance (red circles, D/d=2. 3) tapered section with capacitive compensation (red doted area) • The stripline impedance need to be matched to Z 0 = 50 Ω – Stripline dimensions and shape are critical! – Stripline to coaxial feedthrough transition is very critical!! 20 th September 2016 – Wideband Kicker – Electromagnetic Design (M. Wendt) Page 11

Impedance Optimization wrong feedthrough dimensions too low stripline impedance capacitive feedthrough transition • Impedance matching and mechanical dimensions – go hand in hand Ø stripline width, blends, tapers, feedthrough and other dimensions are critical Ø waveguide pocket and stripline height dimensions are less critical – Define overall performance in terms of bandwidth and amplifier operation 20 th September 2016 – Wideband Kicker – Electromagnetic Design (M. Wendt) Page 12

Summary • Slotted waveguide stripline kicker structure EM analysis – Tr. shunt impedance estimates ~30 kΩ Ø To be verified! – Usable bandwidth DC – ~1 GHz – Some mechanical dimensions and tolerances are critical Ø Evaluated and optimized dimensions and shapes cannot be simplified or altered! • Additional PIC simulation of the final version is underway… • Requires a broadband amplifier – with linear phase (constant group delay) DC – 1 GHz 20 th September 2016 – Wideband Kicker – Electromagnetic Design (M. Wendt) Page 13