140GHz Gyroklystron Design Progress

1
140GHz Gyroklystron Design Progress Colin
Joye Plasma Science and Fusion Center Massachusett
s Institute of Technology Cambridge, MA
02139 July 8, 2004
2
GKL design goals

• Power 100 W CW output from the tube.
• 3dB bandwidth of 1 GHz.
• Current MAGY results at
• 4 perp. velocity spread
• 130 W
• 1.0 GHz of 3dB BW
• 1.3 GHz at 50 watts

3
Block Diagram
4
Why GKL design is so hard

• There are around 30 independent parameters
• Beam voltage, beam current, beam alpha
• Magnetic field, velocity spread value
• Uptaper length and radius
• 5 cavity radii, 5 cavity lengths, 5 cavity ohmic
  Qs 1 diffractive Q
• 4 drift lengths, 1 drift space radius
• Input power and input frequency
• Optimization stumbling blocks
• Oversaturation too much gain over a particular
  frequency range
• Frequency pulling which cavity is the cause?
• Each parameter contributes something to the whole
  spectrum

5
GKL Design Parameters

• 5 Cavities
• V0 15 kV
• I0 0.15 A
• a 1.5
• Perp. velocity spread 4
• Input power 25mW
• B0 51.4 0.1 kG
• Now lets see some graphs!............

6
Cavity shape
7
Output cavity
Cold Cavity Code predicts a nice waveform f0
140.534 GHz Qdiff 358
8
Input power variation (1/9)
Input power variations show good
behavior. Design values give 130 W peak 1.0
GHz 3dB BW 1.3 GHz at 50W
9
Gain vs. input power (2/9)
The gain shows good behavior at all input
powers. Linear gain is 39dB Design value gain is
37dB
10
Beam current variation (3/9)
Power can be increased using I0 at the expense of
3dB BW. BW _at_ 50W goes up, though!
11
Pitch factor variation (4/9)
Pitch factor is sensitive, but can be compensated
by I0. BW _at_ 50W decreases with decreasing a.
12
v? spread variation (5/9)
This design is robust up to 6 perp vel sprd.
13
Beam voltage variation (6/9)
As predicted by linear theory, the gain generally
decreases with increasing voltage. Low voltage
is good for gain but bad for space charge.
14
Magnetic field variation (7/9)

• Even over a B0 deviation of 0.4, over 1 GHz of
  BW is available at 50W.
• Magnexs spec of 0.1ppm/hr Adjust the magnet
  every 5 years!

15
Electric field profile (8/9)
The field profile looks fairly clean. The output
profile will change slightly after a non-linear
taper is designed.
16
Short pulse shape (9/9)
The short pulse behavior is quite good. Pulse
turn-on and turn-off delays 1ns
17
Tuning sensitivity

• The cavities need to be tuned to within 50 to 100
  MHz.
• Electroforming gives a tolerance of 0.2 mils, or
  300 MHz.
• We need a way to accurately tune the structure
  a 140 GHz VNA (hint hint) and maybe a
  Haimson-approved premium gold-plated hammer.

18
Conclusion

• The present GKL circuit design seems solid.
• Still to do
• Nonlinear uptaper to minimize mode conversion
• MIG design
• Mode converter (TE02 to TE01) and input coupler
• Purchase RF source
• Evaluate PBG?