Narrow waveguide experiments

Kazue Yokoyama, Toshiyasu Higo, Yasuo Higashi, Noboru Kudo, Shuji Matsumoto, Shigeki Fukuda, Mitsuo Akemoto,Mitsuhiro Yoshida, Tetsuo Shidara, Hiromitsu NakajimaAccelerator Laboratory, KEK

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.

3/18

Contents1. Design and Fabrication of Narrow Waveguide2. Experimental Setup CU002 at XTF (old X-band Test Facility at KEK) SUS003 at Nextef (New X-band Test Facility at KEK) Scheme of RF Processing

3. Results of High-Power Testing 4. Measurement system of BDR5. Conclusion and Summary

4/18

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

5/18

Fabrication

4 pieces for a narrow waveguide braze bonding in hydrogen furnace

Cu-002 SUS-003 Cu-004

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

VSWR@11.424 GHz

1.4 1.12 1.02

status PrototypeTested at XTF

Under Testing at Nextef

plan to be tested on June.

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

6/18

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

7/18

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

High powerDummy 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)

8/18

Processing Scheme @XTF

Interlock system -> a reflect power is large (vswr>1.4) and vacuum degrades for RF component protection.

Options: Controlled fixed time and power step RF breakdowns caused deterioration of vacuum Daytime processing

Power ~25 MW

Time (~8h)

9/18

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

10/18

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.

@XTF @Nextef

11/18

Accumulated number of breakdown events vs. P*T^0.5 during processing

P*T^0.5 – the product of RF power and the square root of the pulse width

The temperature related parameter, P*T1/2, attained approximately 400 MWns1/2 of Cu002 and 900 MWns1/2 of SUS003.

@XTF

@Nextef

12/18

Results of Cu-002 and SUS-003

SUS003 attained higher electric field than Cu002.

13/18

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.

@Nextef@XTF

14/18

BD Measurement System

A rf pulse is detected with a crystal diode and a OSC that calculates a power, vswr and power loss.

After processing, we’ve measured the rate of breakdown events at a constant power for one day at 50 pps.

oscilloscope

15/18

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

16/18

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

17/18

Observation of Breakdown surface (top) by SEM and Laser Microscope

27.70 m

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.

19/18

Thank you for your attention. !

20/18

SummaryCu-002 SUS-003 Cu-004

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

VSWR @11.424 GHz 1.4 1.12 1.02

status PrototypeTested at XTF

Under Testing at Nextef

next month

Power @ 50 - 400ns [MW]

40 - 20 < 55? – 45

Field @ 50 -400ns [MV/m]

140 - 100 < 140? - 120

P*T1/2 [MWns1/2] 400 900

BDR × measured now

Observation of surface SEM, Laser Microscope

next month?

21/18

Breakdown location

Acoustic sensors

Area of frequent breakdowns

PMT6 PMT5

Cu-002

?