Chris Gilmour Studies into the Design of a Higher Efficiency Ku Band ring-loop Travelling Wave Tube SWS using the CST PIC Software.... the power in microwaves!
History TMD have been making ring-loop TWTs since 1990 Continuous development: Customers want higher average power and higher efficiency Competitors are producing higher efficiency tubes Studies using ESA TWA_2D large signal gain program showed higher beam efficiency in an X band TWT by employing new style of velocity taper The predictions were verified by test results on 2 TWT s This study aims at increasing efficiency in a lower peak power Ku band TWT using CST PIC code
Problems with other software ESA software only to be used for non-military applications Some numbers have to be fiddled to get meaningful answers Synchronism seemed to occur at incorrect cathode voltages Interpretation of output power results required as power seems to vacillate above saturated level Although peak power was increased significantly in X band, the predicted increases of 2 db were nowhere near realised.
Model of 3 Pitch Output SWS with Buncher Model fully parameterised Matched r.f. output r.f. input to model Dummy r.f. output Matched dummy input e- beam Buncher section SWS made of 3 regions of differing pitch identified by colour
Zoom in on slow-wave Structure Loops Stop for support bars In model as will absorb some forward and some secondary electrons However tuner makes this redundant Rings 3 dielectric support bars
Output ( and Input ) Match Tuner On a real tube the coaxial output usually feeds into waveguide Output coaxial outer Output coaxial pin Slow wave structure Artificial Tuner On a real tube the match is usually tuned externally In the waveguide
Dispersion Comparison with Alternative but Previously Proven Software Dispersion Curve Vector Fields v CST Eigenmode Frequency (GHz) 20 19 18 17 16 15 14 13 12 10 11 9 8 7 6 5 4 3 2 1 0 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 Phase shift per pitch (pi radians) Vector Fields CST sws/10 CST sws/15 CST sws/5 CST sws/7
Meshing (sws inner radius/10)
Dispersion Determination Using Eigenmode Solver 3pi radians phase shift over 10 sws pitches
Input Match Graph
r.f transmission (no electrons) (all pitches identical at 0.486mm)
Beam Bunching
Build Up Of Output Power Over Time With Electron Beam
Slater s Saturated Output Power Slater's Predictions for Saturated Output Power (for standard pitch with different loss rates on sws) Peak Power (watts) 600 550 500 450 400 350 300 250 200 150 100 50 0 15 15.25 15.5 15.75 16 16.25 16.5 16.75 17 17.25 Frequency (GHz) 0.5 db/iinch 1 db/inch 2 db/inch 2 db/inch 82% beam size
CST Power Transfer Data for Various Output Configurations CST Output Model of rf Transfer 8kV 350mA 0.425mm beam Output Power (watts) 650 600 550 500 450 400 350 300 250 200 150 100 50 0 46 47 48 49 50 51 52 53 54 55 rf input power to artificial buncher section of model (dbm) new style 15.3 GHz new style 16.0 GHz new style 16.3 GHz new style 16.5 Ghz existing taper 15.0 GHz existing taper15.3 GHz existing taper 15.7 GHz existing taper 16 GHz existing taper 16.5 GHz no taper 16 GHz no taper 15.3 GHz No taper 15.7Ghz no taper 16.3 GHz no taper 16.5 GHz
Output Power Predicted by TWT_2D ESA TWA_2D Saturated Power Peak Power (watts) 650 600 550 500 450 400 350 300 250 200 150 100 50 0 14.5 15 15.5 16 16.5 Frequency (GHz) new style standard
CST Saturated Power CST Output Model of Saturated Output Power 8 kv 350 ma 0.425mm beam Saturated Power (W) 650 600 550 500 450 400 350 300 250 200 150 100 50 0 15 15.2 15.4 15.6 15.8 16 16.2 16.4 16.6 Frequency GHz No taper Existing taper new style Note existing tapered power about 60% of max at 1GHz above frequency of max
Experimental Saturated Output Power Experimental Saturated Power 8 kv 350 ma 450 400 350 Peak Power (W) 300 250 200 150 100 50 0 15 15.2 15.4 15.6 15.8 16 16.2 16.4 16.6 Frequency (GHz) Allowing for match Measured Note power about 60% of max at 1GHz above frequency of max
Phase Information on Beam
Phase Information on Beam
Electron Interception Information
Beam with 5 degree angular spread
Summary of Results Good agreement of dispersion characteristics with other software and practical measurements Good agreement of form of output power versus frequency with other software, analytical theory and practical measurements However predicted power much higher than measured This is also true for other frequency/power level TWTs Predicted power levels little reduced by increasing loss in model, reducing beam size and introducing non-laminar beam
Model with 3 Stage Depressed Collector (also parameterised)
Beam into a 3 Stage Depressed Collector (no secondaries)
Beam into a Depressed Collector (with secondaries)
Beam Sequence into a Depressed Collector (with secondaries)
Summary Using CST we are now able to: Optimise the input and output match. Characterise the beam bunching and interaction. Optimise the slow wave structure to gain maximum output power.
Future work With CST we will be able to look at: Further characterising the effects of secondary electrons from the collector, and reducing them. Optimising multi-stage collectors for increased efficiency. Introducing losses onto the slow wave structure. Create a beam with the non-laminar characteristics of the beam from a gridded electron gun.
Thankyou