Transcript Practical power capability of rectangular waveguides

Practical power capability of rectangular
waveguides
Example: Waveguide for 1 GHz rf system for the CLIC
drive beam
Main Source of inspiration:
Diploma thesis of Sebastian Göller, DESY
“Investigation of high power limitation of waveguide
elements at FLASH”
Steffen Doebert, BE-RF-MK
Requirements for the 1 GHz rf system for the CLIC drive beam
General specifications:
Frequency:
Max. Power:
VSWR:
Waveguide:
Environment:
Flange:
999.5 MHz +/- 50 MHz
20 MW, 150 ms, 50 Hz
1.15:1 max
WR-975, Aluminum
2-3 bar Nitrogen
rectangular flat (parker seal)
WR975: a= 247.65 mm, b =123.8 mm, l= 299.79 mm, lc = 495.9 mm, Zo = 377 W
Theoretical cw power limit for waveguides
Breakdown voltage in air: 30 kV/cm = 3 MV/m
P theo (WR975) = 145 MW
Theoretical cw power limit for waveguides
Theoretical attenuation for waveguides
Additional influencing factors
Average power:
(mainly through heating, see temperature)
Pulse length: (pulse length, rep rate dependence)
Waveguide pressure: (higher is better)
Humidity: (wet is better)
Gas species: (air vs SF6)
VSWR=Vmax/Vmin=(Vf+Vr)/(Vf-Vr), short circuit  Pmax/4
Emax= E0 {2*VSWR/(VSWR+1)}2
Geometrical field enhancement: waveguide components, HOM’s
Temperature: (hot is worse)
Thesis at DESY based on field emission and experiments !
Theoretical cw power limit for waveguides for different gases
and pressure
Ionization rate vs other electron loss mechanisms
Temperature dependence: p= N/V kbT
(for constant p, density goes down with higher T)
Theoretical cw power limit for waveguides for different gases
and pressure
Humidity
(water vapour has a electronegative effect)
Quite difficult in practice to control, other effects condensation, rust
Roughness or FN-beta
Conclusion typical beta ~ 2
Geometrical field enhancement
Higher Order Modes
could as well contribute
(Study concludes not an
issue for DESY
parameters)
Experimental set up to verify assumptions
Pulse length and rep rate dependence
DESY decided to treat their case as cw
Practical power limit as suggested by DESY
f: field enhancement (1= straight, 1.5 E-bend, 2 bellows)
b: FN-beta (typically 2 in waveguides)
h: humidity (h=1)
p: absolute pressure in bars
T: temperature (important for high average power)
VSWR: voltage standing wave ratio (typical value of components
1.2)
Therefore with typical values: P = Ptheo *0.093
For WR650: Ptheo = 58 MW, Ppractical = 5.4 MW in good agreement
with experiments and experience
Predicted power limits for the 1 GHz system
P theo (WR975) = 145 MW
P practical = 145 *0.093 = 13.5 MW
P (2 bar) = 34 MW
P (3 bar) = 58 MW
Keep in mind this assumes CW operation
(factor 1.5-2 according to DESY assumptions)
Mega Industries typical specs:
Pmax/min
(MW)
Conclusion
With 2 bars dry air and carefully chosen waveguide components,
20 MW peak power should be save to operate
Additional safety margin can come from using higher pressure
or SF6 if needed
Status of the L-band klystron:
Call for tender preparation:
Parameters, 20 MW, 150 ms, 50 Hz, ~70% efficiency
News from CTF3
Detailed talk by Robert during the next project meeting,
24th of May
 TBL operation with up to 21 A, 720 MW total, 40 MeV
deceleration average (33%, peak close to 40%)
 TBTS, not much time yet with factor 8, Thursday and Friday
Recheck wake field monitors at higher power, some noise
 BPM’s, CLIC prototype BPMs for main beam and drive beam
successfully tested, behaviour as expected
 Successful tests of phase monitors for phase feed forward
News from CTF3
News from CTF3
News from CTF3