A 94 GHz Overmoded Traveling Wave Tube (TWT) Amplifier

Transcription

1 1 A 94 GHz Overmoded Traveling Wave Tube (TWT) Amplifier Elizabeth J. Kowalski MIT Plasma Science and Fusion Center MURI Teleseminar December 5, 2014

2 2 Outline Introduction TWT Design and Cold Tests TWT Experiment Summary

3 3 TWT Operation Magnet (Solenoid) Electron Gun Slow wave structure Beam tunnel RF in Pass electrons through structure with v e =v p (synchronous operation) Electrons bunch longitudinally, causing exponential growth in the field until saturation RF out Bunching Accelerating Field Electric field in TWT: Exponential growth saturation Decelerating Field

4 E TWT Gain Theory Coupled dispersion relation via circuit theory and electronic equations gives relation of propagation constants: Including loss (d), space charge effects (QC), and non-synchronous operation (b): β = β e + ε Growing Electric Field Solution: Gain (in db) is calculated to be: Citation: Tsimring, 2007 Distance 4

5 Overmoded TWT Project Goals * overmoded TWT design Overmoded design can push to high power at high frequency Our goal is to experimentally validate a 94 GHz overmoded coupled cavity TWT with 30 db of gain (300 W peak power) Low magnetic field (2.5 kg solenoid magnet) Single-beam Pierce electron gun Simple manufacturing of cavity 5

6 6 Outline Introduction TWT Design and Cold Tests TWT Experiment Summary

7 TWT Experimental Overview Solenoid Magnet, 26 cm RF Windows Collector to pump Electron Gun Cavity features highlighted in cyan Operation Parameters Frequency V 0 Current Beam Radius Beam Tunnel Radius Cavities 87 Interaction Length Magnetic Field 94 GHz 31.1 kv 310 ma 0.3 mm 0.4 mm 6.88 cm 2.5 kg Pulse length 2 µs 7

8 2.54 mm Overmoded Cavity Design TM 31 operation at 94 GHz allows for: larger beam tunnel and smaller magnetic field requirements than fundamental mode designs easier manufacturing (larger cavities) RF out 5.8 mm 0.8 mm 0.8 mm TM 31 Electric Field Magnitude RF in beam tunnel (purple indicates vacuum) beam tunnel 8

9 9 Lower Order Modes TM 31 operation allows for a larger cavity design than fundamental mode operation Electric Field Magnitude TM31 TM21 TM11

10 Lossy Dielectric Loading Dielectric placement chosen such that TM 31 is not affected while damping the TM 11 and TM 21 modes AlN composite with tanδ = 0.25 HFSS Simulation of Dielectric Loading TM 11 TM 31 Electric Field Magnitude 0.6 mm TM11 TM21 TM31 TM mm Dielectric 10

11 11 1-D and 3-D Simulations electron beam RF in RF out Analytical calculations and 1-D LMSuite Latte simulations performed with coupling impedance and dispersion relation calculated from HFSS 87-Cavity structure simulated in CST Particle Studio (3-D PIC) with dielectric loading Experimentally anticipated electron beam and magnetic field parameters were used

12 12 TWT Expected Performance CST Particle Studio (3D PIC code) and 1-D LMSuite Latte simulations Structure Parameters Cavities 87 Length 6.88 cm Coupling Impedance 2.8 Ohms TM 31 Loss 4 db/cm Simulation Results Center Frequency GHz Gain 32 db Peak Power 300 W Bandwidth 200 MHz

13 Cold Test Manufacturing CNC milling with inserts for dielectric placement Different materials tested: OFHC copper, AL60 and AL25 Glidcop dielectric placement 19 cavities; OFHC WR10 9 cavities; Glidcop WR mm Beam Tunnel 13

14 14 Cold Test Measurements Good agreement with theory Higher transmission in glidcop structure Effective suppression of modes AlN composite dielectric: ε r = 25 tanδ = 0.24 Comparison to theory (9-cavity glidcop, no diel) TM31 TM41 Dielectric Losses (19-cavity glidcop) TM31 TM41

15 15 Outline Introduction TWT Design and Cold Tests TWT Experiment Summary

16 16 94 GHz TWT Experiment Set Up 94 GHz EIK (driving source) 2.5 kg Solenoid Magnet 30 kv Electron Gun 2 microsecond pulse modulator

17 17 Solenoid Magnet Tested Solenoid magnet with 2.5 kg field for 10 cm tested with iron pole piece Pole piece designed for high-compression electron gun Simulations with Poisson 3-axis measurements showed 2-3 mm misalignment between magnetic field axis and bore axis; iron pole pieces adjusted to compensate 1.8 kg Measurement

18 18 Pierce electron gun designed with Michelle for operation at: 310 ma at 30 kv, 0.3 mm beam radius Current density at cathode = 4 A/cm 2 ; 1.63 mm/0.25mm compression ratio Electron Gun Design ~ 6 in. Ceramic Cathode Anode Pole piece Michelle Simulation

19 19 Electron Gun Manufactured Designed and built at MIT Cathode manufactured by Heatwave Labs Anode Cathode Beam Test Assembly

20 1.1 cm Beam Test Operation Solenoid Magnet, 26 cm RF Windows Collector to pump Electron Gun Operation Parameters Frequency V 0 Current Beam diameter Current Density at cathode Magnetic Field 94 GHz 31.1 kv 310 ma 0.6 mm 4 A/cm kg Pulse length 2 µs 1.4 cm 0.8 mm beam tunnel Electron beam 20

21 21 Electron Gun Beam Test Child-Langmuir curve calculated for P = 0.06 micropervs, where Experiment has been run to 31 kv with 306 ±6 ma at collector (20 ma at anode, 23 ma at beam tunnel) Measurements agree well with expected current

22 22 87-Cavity Structure Assembly 87 Cavities, AL60 LOX glidcop, direct machined Cold tests agreed with theory; full suppression of TM 41 mode

23 Device Gain (db) 23 Results: 21 db device gain Zero-drive stable operation with no evidence of unwanted modes 30.6 kv operation with 250 ma collector current and GHz input signal: 21±2 db linear device gain measured 27 W peak device output power Bandwidth: High gain (30.6 kv) operation point had 30 MHz bandwidth Alternate 28.7 kv operating point had about 110 MHz bandwidth, but much less linear gain Input signal from AMC: solid-state EIO: Klystron

24 24 Results: 27 db circuit gain Circuit gain accounts for coupling losses in the device and is a measure of the gain occurring in the TWT interaction circuit Measured 6±1 db input and output coupling losses 27±2 db circuit linear gain 55 W peak circuit power output Adjusting CST for 250 ma interaction current gives the theoretical circuit gain for the experimental conditions: 28 db linear gain 100 W peak power Input signal from AMC: solid-state EIO: Klystron

25 Summary Designed 94 GHz overmoded coupled cavity TWT Simulated 32 db of gain, 300 W peak power Cold Tests of structures confirm TWT simulations, demonstrate successful manufacturing of cavities and show suppression of unwanted modes with AlN dielectric Electron gun beam test showed 306±6 ma at of current at collector for 31 kv operation and agreed well with theory Overmoded W-band TWT experiment is operational Zero-drive stable with no evidence of fundamental or unwanted modes 27 ±2 db circuit gain (21 ±2 db device gain) at GHz with 250 ma collector current 27 W peak output power at GHz (55 W peak circuit power) Adjusting for experimental conditions, CST predicts 28 db circuit gain, 100 W peak power 25

26 26 Acknowledgements Waves and Beams Division at MIT Plasma Science and Fusion Center Faculty and Staff William Guss Sudheer Jawla Ivan Mastovsky Michael Shapiro Richard Temkin Paul Thomas Paul Woskov Sergey Arsenyev Jason Hummelt Xueying Lu Brian Munroe Grad Students Undergrad Students Samantha Lewis Sam Schaub Alexander Soane Haoran Xu JieXi Zhang Research supported by the Air Force Office of Scientific Research Program on Plasma and Electro-Energetic Physics