325 MHz IQ Modulator

Ding Sun and David Wildman

Fermilab Accelerator Advisory CommitteeMay 10th – 12th , 2006

2Fermilab

Outline

• IQ Modulator• Circulator (status and plan)• Fast Phase Shifter (status and plan)• Hybrid (status and plan)• Recent High Power Test Results• Conclusions

3Fermilab

325 MHz RF Distribution System

Pulse Transformer& Oil Tank

IGBT Switch & Bouncer

CAP

BANK

10 kV110 kV

Charging

Supply

300kW

MODULATOR: FNAL/TTF Reconfigurable for 1,2 or 3 msec beam pulse

SingleJPARC Klystron325MHz

3 MW

WR2300 Distribution Waveguide

TO

SH

IBA

E

3740

A

I

Q

M

I

Q

M

I

Q

M

I

Q

M

R F Q

I

Q

M

Cables inTunnel

Fast Ferrite Isolated I/Q Modulators

RF Couplers

S

I

Q

M

S

I

Q

M

R

I

Q

M

S

I

Q

M

S

I

Q

M

R

I

Q

M

600kW 40 kWmax.

D

I

Q

M

S

I

Q

M

R

I

Q

M

D

I

Q

M

S

I

Q

M

R

I

Q

M

120 kWmax.

10kV

H-

Room TemperatureCopper Cavities (16+2)

Radio FrequencyQuadrupole

Cryomodule #1-2 Single- Spoke

Resonators 1 (18)

Cryomodules#3 -4 Single - Spoke

Resonators 2 (22)

5 kW..

MEBTRT-TSR

… 60 kWmax.

110 MeV

4Fermilab

IQ Modulator (Function)

• The entire front-end linac up to 90 MeV is powered by a single 2.5 MW klystron.

• RF power is carried by a single WR 2300 waveguide alongside the beam line and partially extracted by a waveguide-coax coupler at the location of each RF structure.

• IQ modulators are used to control phase and amplitude of the input power for each RF structure.

• Circulator + hybrid + ferrite phase shifters

5Fermilab

IQ Modulator

Box size: 24” x 20”

6Fermilab

IQ Modulator (Specification)

• Peak Power– 40 kW ~ 120 kW

– 275 kW (for each drive loop of RFQ)

• Tuning Range– Phase: +/- 45 degree

– Amplitude: +/- 1.5 db

• Phase Tuner Slew Rate– 1 degree/1sec

7Fermilab

IQ Modulator (R&D Plan, 2005)

• Design and make individual circulator, hybrid and fast ferrite phase shifter.

• Power test them separately, identify problems and modify design.

• Integrate them and power test.

8Fermilab

Circulator (status)

• Stripline Y-junction Type

• Prototype (low power): RF part finished, final permanent magnet design to be completed.

• A prototype for “high” power test has been designed. 50% of mechanical drawing completed.

~ 8 inch

9Fermilab

Circulator (measured S parameters)

Isolation: -30 db

Insertion lose: -0.18 db

10Fermilab

Circulator (plan)

• Fabricate the prototype circulator (2-3 months) and measure it at low power level, modify it if necessary.

• High power test to determine:

– Power limit for this initial design

– Thermal effect

• Increase height for higher power handling capability.

• How to compensate thermal effect (cooling control).

• Finalize permanent magnet design and mech. detail.

• Time: 12 - 24 months.

• Cost: $ 20 - 30 k (cooling control not included)

11Fermilab

Phase Shifter

• Coax line is filled with ferrite material and shorted at end: power is fully reflected.

• Effective length of coax line is controlled by magnetic field applied with a solenoid.

12Fermilab

Phase Shifter (Status)

• High power tested:– 1 5/8” O.D. coax line: 85 kW

– 3 1/8” O.D. coax line: 440 kW

(2.5” long ferrite)

• Speed of phase change

(low power):

120 degree/50 sec50

A /

div

20 d

egre

e /

div

Blue: current of solenoidPurple: phase of output power

13Fermilab

Phase Shifter (plan)

• Fast phase shift during high power pulse• Measure thermal heating effects at full power • High power test up to 400 kW with 3" O.D. x

5" long garnet • Control phase and amplitude to real cavity

14Fermilab

Hybrid

• A quadrature branch line coupler made of copper sheet with thickness of 0.125”

• A matching section at end of each arm for transition of suspended striplines to 1 5/8” coaxial line ports

• No cooling• Easy fabrication – low cost• Bandwidth: ~60 MHz

Box: 20” x 20”

15Fermilab

Hybrid (measured S parameters)

phase: -89.7 degree

Amplitude: -2.99/-3.18 db

Reflection: ~ -39db

Isolation: ~ -39db

5db/div

FrequencyFrequency

5db/div

1db/div

90

degr

ee/d

iv

Port1 - Port3

Port1 – Port2

16Fermilab

Hybrid (status and plan)

• Status: high power tested: 170 kW without failure under full reflection condition

• Plan: – Add directional coupler within the box for signal

processing (cost saving).

– Further test at FNAL to see power limit using fixed coax line sections with various lengths.

– Modify design (increase height) for higher power (~ 275 kW for RFQ).

17Fermilab

Recent High Power Test Results

• A hybrid and a phase shifter together (a partial IQ Modulator) was tested up to 170 kW input power level (limit of 1 5/8” phase shifter).

• APS 1.3 MW CW Klystron configured in pulsed mode (pulse length: 4 ms, rep rate:1Hz).

• Port #2 is connected with a 1 5/8” ferrite phase shifter, port #3 is shorted.

• Amplitude and phase of output power (port #4) was controlled at all power levels by adjusting phase shifter.

18Fermilab

Recent High Power Test (setup)

19Fermilab

Recent High Power Test Results (plot)

Amplitude Modulation of Oupt Power at Various Input Power Level

0

20

40

60

80

100

120

140

160

180

-60 -40 -20 0 20 40 60

Phase difference (degree) between port #2 and #3

Am

pli

tud

e o

f O

utp

ut

po

wer

at

po

rt #

4

20 kW

30 kW

40 kW

50 kW

70 kW

80 kW

90 kW

100 kW

120 kW

140 kW

160 kW

Phase Modulation of Output Power at Various Input Power Level

0

20

40

60

80

100

120

140

160

-60 -40 -20 0 20 40 60

Phase difference (degree) between port #2 and #3

Ph

ase

of

ou

tpu

t p

ow

er a

t p

ort

#4

20 kW

30 kW

40 kW

50 kW

80 kW

100 kW

120 kW

160 kW

20Fermilab

1.3 GHz IQ Modulator

• For main linac 1.3 GHz superconducting cavities.

• Waveguide version will be delivered in July, 2006.

• Can be used for ILC.

21Fermilab

Conclusions

• IQ Modulator R&D are doing well following original plan.

• Hybrid and ferrite phase shifter together have been (1) power tested up to 170 kW which exceeds all power requirement of front linac except for RFQ and (2) functioned as IQ modulator at high power levels.

• For modest cost so far, IQ Modulator R&D is well worthwhile: enable us to have experience on how to make high power ferrite device work for HINS and possibly ILC.

• We have a plan for what to do next to get to a 325 MHz Front End Linac in Meson.