Recent LFD Control Results from FNAL

Yuriy PischalnikovWarren Schappert

TTF/FLASH 9mA Meeting on Cavity Gradient FlatnessJune 01, 2010

Resonance Control at FNAL

• Recently tested first blade tuner equipped cavity in HTS at FNAL

• Two piezos to control LFD

Measuring LFD

• Complex envelope of the probe signal described by 1st order ODE

• Rearrange terms to extract the width and the dynamic detuning from the probe and forward IF signals

PP

FdtdP

P

FPP

dtdP

P

FPidt

dP

*

2/1*

*

*

2/1

2/12/1

2Im

2Re

Re

2

Correcting the IF Signals• Accurate detuning

calculation requires accurate LLRF signals

• Change the piezo bias and examine IF signals for– Saturation

• Decay of probe and reflected should be exponential

– Cross-contamination of forward and reflected signals

• Amplitude and phase variations of the forward power during the fill and flattop

• Tails on forward power during the decay

100 150 200 250 300 350 400 450

10-3

10-2

10-1

-0.5 0 0.5 1 1.5 2 2.5 3 3.5 40

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

Uncorrected and Corrected Signals

• Applied corrections based on assumption that forward tail is due exclusively to cross-contamination by reflected signal– Assumption is likely

only partially true• Correction reduces

the amplitude and phase variation and forward signal from 5-15% to a few percent

-1 0 1 2 3 410

-4

10-2

100

102

Uncorrected

Time (ms)

For

war

d M

agni

tude

(A

rbitr

ary)

-1 0 1 2 3 410

-4

10-2

100

102

Corrected

Time (ms)

For

war

d M

agni

tude

(A

rbitr

ary)

0 0.2 0.4 0.6 0.8 1 1.2 1.40

5

10

15

20

25

30Uncorrected

Time (ms)

For

war

d M

agni

tude

(A

rbitr

ary)

0 0.2 0.4 0.6 0.8 1 1.2 1.40

5

10

15

20

25

30Corrected

Time (ms)

For

war

d M

agni

tude

(A

rbitr

ary)

0 0.2 0.4 0.6 0.8 1 1.2 1.4-10

-5

0

5

10Uncorrected

Time (ms)

For

war

d P

hase

(D

egre

es)

0 0.2 0.4 0.6 0.8 1 1.2 1.4-10

-5

0

5

10Corrected

Time (ms)

For

war

d P

hase

(D

egre

es)

Delay Scan

• Developed and tested in CCII

• Excite piezo with a series of short pulses at various delays

• Measure the dynamic detuning for each pulse

Response Matrix• Subtract the mean

detuning (LFD) to get response matrix

• Invert the response matrix and determine combination of pulses needed to cancel out the mean using LS

Pulse

Sam

ple

10 20 30 40 50 60 70 80 90 100

20

40

60

80

100

120

140

160

180

-100

-50

0

50

100

0 20 40 60 80 100 120 140 160 180 200-200

-100

0

100

200

300

400

500

600

700

0 20 40 60 80 100 120 140 160 180 2000

100

200

300

400

500

600

Measuring the Cavity Width

• Plot of P*(P-2*F) vs P*dP/dt should give straight line through origin

• Provides a consistency check on correction and detuning calculations -1.6 -1.4 -1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2

-2500

-2000

-1500

-1000

-500

0

500

Re(

P* dP

/dt)

Re(P*(-P+2*F))

Width Estimation

Gradient: 34.2 MV/m

Contaminination: 0.096+-0.033i

Forward Gain: -0.029+-0.398i

Half Width: 219.3

Offset: -1.1 Hz

Bias-18.mat

Dynamic Detuning • Measure the dynamic

detuning at 35 MV/m during the pulse– Using long fill and

short flattop because of HTS klystron limitations

– Each curve is detuning at different piezo bias

• Maximum detuning ~1 kHz during fill and flattop

0 0.5 1 1.5 2 2.5 3 3.5-1000

-500

0

500

1000

1500

2000

2500

Time (ms)

Det

unin

g (H

z)

Lorentz Force Detuning at 34.2 MV/m

Bias-18.mat

Standard Detuning Compensation

• Manual adjustment of pulse parameters– Flat phase

during flattop at 33 MV/m

– Demonstrates that cavity can be tuned at high gradients

LS LFD Compensation• Implemented an adaptive version of

the LS procedure that worked successfully in CCII

• Able to maintain flat phase during both fill and flattop

• Able to track the resonance as cavity was ramped down from 15 MV/m to 35 MV/m and back up again

• Flattop square and phase flat to few degrees at 35 MV/m

• LFD reduced to level of microphonics

Residual Detuning• Detuning at 35 MV/m

after compensation is comparable to microphonics – 50 Hz Peak– 15 Hz RMS

• Probably limited by low pass filter in FF loop– May be possible to

reduce residual detuning further by increasing the LPF cutoff frequency

One Possible Approach• Cavity response described

well by exponential decay• For a given beam current and

Q, can determine the forward current required for a flat field

• Piezo could be used to move cavity off resonance for a short period early in the fill– Provides individual control of

the effective fill time of each cavity

• Optimum detuning profile during fill TBD

Be

FePFillt

t

)1(

)1(/)(

/

BeF

BeF

OffFill

Fill

/

/

Summary

• Excellent results with LFD so far, more progress to come – Reduce LFD from 1kHz (fill+flattop) to <50Hz peak at 35 MV/M

• May be possible to control field flatness of individual cavities by detuning them during the fill– Optimum detuning profile during fill still to be determined

• Piezo is still provides very crude control when compared to LLRF– Response time limited by mechanical resonances of cavity and inertial

of mechanical tuner• Response time may limit to feedforward

– Passive only• Piezo can divert power from cavity to reflected and vice versa but cannot

inject any power into the system