Page 1
High-gradient Experiments with Narrow Waveguides
Kazue Yokoyama, Toshiyasu Higo, Yasuo Higashi, Noboru Kudo, Shuji Matsumoto, Shigeki Fukuda, Mitsuo Akemoto,Mitsuhiro Yoshida, Tetsuo Shidara, Hiromitsu NakajimaAccelerator Laboratory, KEK
WEPP106
Page 2
2/18
Introduction
The study of characteristics of different materials on high-gradient RF breakdown at Nextef (New X-band Test Facility at KEK).
Experiments performed using a narrow waveguide with field of approximately 200 MV/m at RF power of 100 MW.
Status report on prototype copper (#CU002) and stainless-steel (#SUS003) waveguides.
The measurement system for the breakdown rate is set up now.
Page 3
3/18
Rectangular Waveguide: WR90Wavelength converter: Width 22.86(g 32.15 mm) → 14mm (g 76.59 mm)Cosine taper ( 1 g) : Height 10.16 mm → 1 mmCalculated to get a low VSWR by an HFSS.
Narrow Waveguide Design
field gradient of 200 MV/m at an RF power of 100 MW
group velocity of around 0.3c
Page 4
4/18
Fabrication
4 pieces for a narrow waveguide braze bonding in hydrogen furnace
#CU002 #SUS003 #CU004
material OFC SUS316L OFC
anneal 500 C 1020 C 500 C
processing milling, WEDM milling milling
cleaning CP SUSpika CP
bonding Cu/Au/Ni,hydrogen furnace
Cu/Au,hydrogen furnace
Cu/Auhydrogen furnace
[email protected] GHz
1.4 1.12 1.02
Loss [dB] -0.42 -1.56 -0.34
E-Field [MV/m]@1oo MW
212.34 189.25 212.34
status PrototypeTested at XTF
Under Testing at Nextef
plan to be tested
Annealing in a hydrogen furnaceProcessing by milling and and wire electrical discharge machining (WEDM) E-plane finished by milling. 10 m chemical polishing by acid
Page 5
5/18
High powerDummy Load
RF
Narrow waveguide
#CU002 Setup for High-Power processing @XTF
The first high-power test of copper(#CU002) was done at XTF (previous X-band Test Facility at KEK).
Acoustic sensors
PMT 1, 2, 3, 4
PMT 5
Page 6
6/18
#SUS003 Setup for High-Power processing @ Nextef
We are on going high power testing of stainless-steel(#SUS003) at Klystron Test Stand.
PPM Klystron
Narrow waveguidein 5mm lead shield
PMT 5, 6
Acoustic sensors
Dummy Load
PMT 1, 2, 3, 4
#CU002 tested at XTF(06.11 ~ 07.01) Moving to Nextef( ~ 07.04) Using for system ch
ecking ( ~ 07.05) #SUS003 Tested at Nextef ( ~ now)
Page 7
7/18
Processing Scheme @XTF
Interlock system -> a reflect power is large (vswr>1.4) and vacuum degrades for RF component protection.
Options: Controled fixed time and power step RF breakdowns caused deterioration of vacuum Daytime processing
Power ~25 MW
Time (~8h)
Page 8
8/18
Control time step (flexible) Control power step by limiting
Vac. (flexible) Processing time is almost 24 H.
We’re seeking for ways of processing.
Processing Scheme @Nextef
Vac.< 1E-6 Pa
Power~30MW
Power step
Vac.
Time (4h)
Pulse width 200ns -> 400ns
Page 9
9/18
Processing history (Accumulated No. breakdown events vs. power
RF pulse ranged 50ns to 400ns with 50MW at repetition rate of 50pps. #CU002 had more breakdown events than #SUS003.
Page 10
10/18
Results of #CU002 and #SUS003
#SUS003 attained higher electric field than #CU002. The temperature related parameter, P*T1/2, attained approximately
400 MWns1/2 of #CU002 and 900 MWns1/2 of #SUS003.
P*T^0.5 – the product of RF power and the square root of the pulse width
Page 11
11/18
RF Power vs. number of BD events
Many breakdown events at pulse width > 100ns and power 20MW.
Few RF break down events at 50 ns and 100ns .
We had a guard window problem around 200ns.
Page 12
12/18
A rf pulse is detected with a crystal diode and a OSC that calculates a power, vswr and power loss.
oscilloscopeBD Measurement System
Page 13
13/18
Data taking from OSC 10 pulses at a Syste
m down (HV, trigger and rf off).
Changed pulses with 5 % form a normal pulse (area and peak).
Estimated power, vswr and power loss
We are testing this system now.
Problem is pulse shape is unstable from pulse to pulse (rf jitter, rf power and noise).
Typical rf pulses at a breakdown events.
Transmitted
Forward
Reflected
Reflected upper stream
Page 14
14/18
After processing, we’ve measured the rate of breakdown events (BDR) at a constant power for one day at 50Hz.
BDR of #SUS003
Page 15
15/18
#CU002 After high-power processing
top
body
Many breakdown damages were seen on the E-plane surface.
The surface is intensively damaged, and it could also melt due to breakdown.
Observation area
Page 16
16/18
Observation of Breakdown surface (top) by SEM and Laser Microscope
27.70 m
Page 18
18/18
Conclusion RF breakdown studies on different material has just
begun. Prototype #CU002 and #SUS003 had been tested
under different system conditions. Number of break down events for #SUS003 is less
than that for #CU002 which may be a result of different systems .
We’re testing a processing scheme. We’re going to observe the surface of #SUS003
after measuring BDR. We’re going to test #CU004, other stainless-steel
waveguides and other materials.