Design of a Novel 140 GHz,

1
Design of a Novel 140 GHz, 1 kW
Gyro-Amplifier Colin D. Joye, A. Cerfon, M. A.
Shapiro, J. R. Sirigiri, R. J. Temkin and A. C.
Torrezan Plasma Science and Fusion
Center Massachusetts Institute of
Technology Cambridge, MA 02139 April 27,
2006 Session 21.4
2
Outline of Presentation

• Introduction, Specifications.
• Confocal waveguide.
• Simulations (linear, nonlinear, backward wave).
• Novel quasi-optical severs.
• Summary

3
Introduction

• Need for high gain, moderate power amplifiers at
  gt100 GHz
• Short pulse spectroscopy
• Electron Paramagnetic Resonance Nanosecond-scale
  pulse sequences.
• Dynamic Nuclear Polarization enhanced Nuclear
  Magnetic Resonance.
• Emerging medical applications.
• Explore a novel confocal structure
• Distributed diffractive loss to stabilize against
  oscillations.
• Avoids use of lossy ceramics or other custom
  absorbers.
• Based on a successful experiment at MIT1.
• Novel high order mode operation

1J. R. Sirigiri, M. A. Shapiro, R. J. Temkin,
High-power 140-GHz quasioptical gyrotron
traveling-wave amplifier, Phys Rev Lett, vol.
90, iss. 25, pp. 258302-4, 2003.
4
Amplifier Diagram
5
Input Transmission Line

• 12.6 mm (0.5 inch) diameter overmoded TE11
  transmission line
• 3.6 0.5 dB loss in 138-142 GHz band

6
Design Parameters
7
Superconducting Magnet

• 6.2T superconducting magnet from Magnex
  Scientific.
• Demonstrated to meet spec.
• 28cm flat field (/-0.5).
• 127mm warm bore diameter.
• 1m bore length.
• Shielded design reduces stray magnetic field
  outside the magnet.

8
Varian VUW-8140 Triode

• Varian VUW-8140 Triode MIG
• V0 30 kV
• Cathode Radius 0.91 cm
• Magnetic compression 25
• Transverse spread lt 4

9
Confocal Waveguide

• Dispersion Relation
• Loss rates
• HFSS

10
Quasi-optical Confocal Structure

• Gaussian-like higher order mode
• Mode selectivity enhanced by diffraction
• Mode spectrum is sparser than a cylindrical
  waveguide circuit
• Suitable for harmonic operation
• Mechanical tuning capability in vacuum
• Surprisingly robust to misalignment
• Diffractive loss avoids custom lossy materials.
• Easily increase loss by cutting down the
  aperture.
• Simple sever implementations

Rc Mirror curvature and separation. a Aperture
size.
11
HFSS Model

• Simple input coupler design
• 5.5 GHz BW
• S11 lt -15 dB

12
Dispersion Relation

• HEMN Modes are spaced equidistant in frequency
  and linearly in M and N,
• The loss rates can be calculated

HE06 -1 dB/cm HE15 -28 dB/cm HE05 -2 dB/cm
Confocal waveguide can be highly selective.
HE05 is the primary BWO mode.
13
Theory vs. HFSS

• The simple Gaussian beam theory matches HFSS
  within a few tenths of a dB/cm.

14
Simulations

• Linear
• Backward wave
• Nonlinear

15
Linear Gain Rate
Need 3 dB/cm to get gt50 dB total.
V0 30kV I0 2A a 0.75 v? spread 0 B0
4.94 T Mode HE06 Rc Lt 6.9mm Rb 1.85mm
Linear theory predicts significant gain as
required to meet spec.
16
Backward Wave Oscillations

• BWO criterion for the HE05 mode using general
  start current equation1
• f 120.2 GHz
• V0 30 kV
• a 0.75
• B0 4.94 T
• Rc 6.9 mm
• Rb 1.85 mm (2nd max)
• For I02A, a3.0mm, we have 7.2 cm section
  maximum lengths.
• (code does not include kzI yet for confocal)

1Nusinovich and Li, Theory of gyro-travelling-wav
e tubes at cyclotron harmonics, Int. J. Elec.,
v. 72 (5-6), pp. 895-907, 1992.
17
Nonlinear Simulation
Sev
Gain profile at 140 GHz

• v? spread 6
• B0 4.97 T (99.7 grazing)
• Rc Lt 6.85mm
• a 3.0 mm
• Pin 10 mW
• Pout 2 kW pk gt 1 kW
• Sat. Gain 53 dB gt 50 dB
• BW(-3dB) 2.4 GHz gt 1GHz

18
Sever Options

• Confocal waveguide allows severs to be
  constructed simply.
• Diffractive sever region of very small aperture.
• HFSS 30 dB loss over 3cm.
• Quasi-optical sever due to the high bounce angle
  of 75o, only a small gap is needed to achieve
  high loss.
• HFSS Undergoing simulation.

Waveguide top view
Waveguide side view
19
Summary

• The design of a novel high-order mode gyro-TWT
  employing diffractive loss via confocal waveguide
  has been presented.
• A simple theory for confocal waveguide was
  discussed.
• Linear gain, backward wave thresholds were shown.
• Nonlinear simulations Requirements are within
  reach.
• Concepts for quasi-optical severs look very
  promising.
• Plan to run first experiment in August, 2006.
• This work is supported by NIH / NIBIB, grant
  R01-EB001965.