Investigation of Radio Frequency Breakdown in Fusion Experiments T.P. Graves, S.J. Wukitch, I.H. Hutchinson MIT Plasma Science and Fusion Center APS-DPP October 2003 Albuquerque, NM
Outline Multipactor Basics Simple multipactor description Multipactor discharge properties Other types of multipactor discharges RF Breakdown Experiment Motivation behind experiment How multipactor discharges apply Experimental Results How this could apply to Fusion/Space systems Future Plans
Abstract An electron multipactor discharge is a resonant vacuum discharge which occurs in RF systems in the MHz to tens of GHz frequency range. This discharge can develop in systems such as microwave devices, RF satellite payloads, or accelerator structures. The RF Breakdown Experiment for Alcator-CMOD, designed to test high-voltage, RF arcing, appears to be limited by coaxial multipactoring within the vacuum system. The underlying physics and properties of multipactor discharges along with the experimental results of the RF Breakdown Experiment will be presented.
Multipactor Basics A multipactor discharge is a resonant condition for electrons Radio Frequency effect MHz to 10 s GHz frequencies Observed in: Accelerators Microwave devices and resonators RF satellite payloads Vacuum conditions required Electron multiplication from secondary electrons Need SEC > 1 and sufficient impact energy Copper SEC δ = 1.3, Energy = 600eV (peak), 200eV (min) (Handbook of Chemistry and Physics, 72 nd Edition) System geometry and E field structure R.A. Kishek, Phys. Plasmas 5, 2120 (1998)
Parallel Plate Multipactor T(0 to π) T(π to 2π ) T(2π to 3π) e - Secondary Electrons E 1 E 2 E 3 Secondary Electrons E max 0 E field wave E min
Multipactor Properties Parallel Plate (PP) Geometry Lorentz equation can be solved for multipactor condition: 4 m V = π ( f d ) 2 e Due to simple geometry, PP situation well studied More complicated due to secondary emission energy range Susceptibility Curve R.A. Kishek, Phys. Plasmas 5, 2120 (1998)
Other Multipactor Characteristics Advantages to Multipactor Discharges Surface conditioning Electron source (Farnsworth Oscillator) Plasma display technology Disadvantages Detunes microwave circuits, dissipating all excess power Typically impossible to push thru discharge Introduces large noise signals in space applications Surface heating Vacuum window breakage Induce vacuum breakdown (arcing)
Preventing Multipactoring Three General Approaches Surface conditioning Increases the necessary energy for secondary emission Low SEC Coatings Less secondary electrons, yet tend to wear off with time Change geometry Make geometry such that multipactoring is impossible
Other Multipactor Regimes Single Surface Dielectric Multipactor E dc V o E rf sin(ωt) + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + Coaxial Transmission Line Multipactor Outer diameter Inner diameter E rf = E o sin( ω t) r
Coaxial Multipactoring Not well explored Space and Fusion applications have vacuum coaxial transmission lines Coaxial lines typically filled with insulator like SF 6 preventing multipactoring Result of non-linear behavior of electrons Single or double surface multipactoring Still need large enough impact energy for secondary emission Other Non-linear Multipactor Regimes Possible Other non-linear, non-uniform field geometries Influence of external magnetic fields in particle orbits R. Woo, J. Appl. Phys. 39, 1528 (1968)
RF Breakdown Experiment Motivation behind experiment Alcator C-MOD utilizes high power (MW) RF systems for ICRH (~80MHz) Empirically determined E-field breakdown limit, E=15kV/cm, on C-MOD Similar limits seen on experiments such as JET and NSTX RF Breakdown Testbed was built to explore E-field limit High Q (~1500) resonator to build up high power (extremely sensitive to small changes) Transition to strip line at high voltage point to locate arcing Problems arose when circulating power levels were limited at surprisingly low voltage (~100V)
RF Breakdown Setup Source 4kW Pulsed ~ DC1 DC2 Double Stub Tuner Vacuum System Base pressure 10-7 torr, He Tested with and without Ground plane (suspected series arcs, redesign)
Amplifier Vacuum Break Directional Couplers Stub #1 Directional Couplers Stub #2
RF Short Vacuum Chamber Vacuum Break
Initial Power Limit Data Limit detected by rise in both reflected power to source (mismatch) and ion gauges (discharge cleaning surfaces) Seems to line up with parallel plate d=1.25 cm multipactor Problem no 1.25 cm gaps in system! Possible strange regime in square vacuum chamber? Possible Coaxial Multipactoring? Next step Achieve better vacuum (10-7 ) to increase necessary impact energy for secondary emission 8-15-03, No Ground Plane Present Base Pressure = 1e-5 torr (prebake)
Why the limit? Baked system 10-7 torr much less adsorbed gases Result: Small increase in limit voltage, but limit now completely independent of pressure Put ground plane back into system Changed voltage pattern completely inside box Result: Very small variation with/without ground plane Replaced quartz window with SS conflat checking for possible dielectric multipactoring Result: No variation with and without window Voltage in vacuum box not constant with frequency Seems that multipactor is not focused in the box
72.25 in 150 MHz 80 MHz Frequencies 78.1975 80.1188 115.25 151.0437 192.8562 239.125 150 MHz 80 MHz CST Fieldscode,T.Graves, P. Koert
Coaxial Multipactor Regimes Each figure below corresponds to the experimentally determined voltage @ that frequency E=8eV E=8eV E=22eV E=23eV J. Irby, T. Graves
Current Results Summary
Application to Fusion Systems Use of vacuum coaxial transmission lines J-port has 2 sections of 4 inch vacuum coaxial lines (as well as poor operation) Situation in plasma non-vacuum operation transmission of power Finite VSWR Can the voltage get low enough and the phasing be just right to initiate a multipactor in the proper region? ITER plans to have multi-wavelength long vacuum coaxial transmission lines for ICRH system
Conclusions Multipactor discharges are resonant electron discharges which effect many different RF systems Typical scaling is V ~ (f d) 2 and can occur in many different geometries including coaxial transmission line The RF Breakdown Experiment seems to be limited by what appears to be a coaxial multipactor and it is this might relate to the much higher voltage limit on C-MOD of 15kV/cm New information is needed to address the problem of coaxial multipactoring
Future Plans Is power limit definitely a multipactor effect? What is the electron impact energy? What mode (harmonic) is the multipactor? Where in the line is it occurring (high voltage point?) Move short outside vacuum region and determine location of multipactor Diagnostic inside coax to investigate multipactor electron potential and energy What happens when there is a net power out? Can you get multipactoring with a non-infinite VSWR - Attach load to resonator to look at VSWR scenario