HOM Experiences at the SNS SCL

HOM Experiences at the SNS SCL

SPL HOM WorkshopCERN

June 26, 2009

Sang-ho Kim

SNS SCL Area Manager

SNS/ORNL

2 Managed by UT-Battellefor the U.S. Department of Energy SPL HOM Workshop at CERN, June 25-26, 2009

Outline

SNS Intro

HOM concerns in SNS SRF cavity Reminder of past histories

HOM coupler issue

Re-evaluation HOM frequency statistics SNS beam Beam induced signal

Summary


SNS Accelerator Complex

945 ns

1 ms macropulse

Cur

rent

mini-pulse

Cur

rent

1ms

Front-End:Produce a 1-msec

long, chopped, H- beam

1 GeV LINAC

Accumulator Ring: Compress 1 msec

long pulse to 700 nsec

Chopper system makes gaps

Ion Source2.5 MeV 1000 MeV87 MeV

CCLCCL SRF, =0.61SRF, =0.61 SRF, =0.81SRF, =0.81

186 MeV 387 MeV

DTLDTLRFQRFQ

Accumulator Ring

RTBT

Target

HEBT

Injection Extraction

RF

Collimators

Liquid Hg Target


SNS SRF cavity

Fundamental Power Coupler

HOM CouplerHOMB

HOM CouplerHOMA

Field Probe

Tuner

Helium Vessel

Major Specifications:Ea=15.9 MV/m at =0.81

Ea=10.2 MV/m at =0.61

&Qo> 5E9 at 2.1 K


Designed to operate at 2.1 K (superfluid helium)

Fundamental power couplers

Return end can

Helium vessels

Space frame

Supply end can

SNS Cryomodule

11*Medium beta

12*High beta

81 cavities


Beam power history

Summer 2007

Winter 2007

Summer 2008

Spring 2007Fall 2006

Winter 2008

Summer2009


Parameters DesignIndividually

achieved

Highest production

beam

Beam Energy (GeV) 1.0 1.01 0.93

Peak Beam current (mA) 38 40 38

Average Beam Current (mA) 26 24 24

Beam Pulse Length (s) 1000 1000 670

Repetition Rate (Hz) 60 60 60

Beam Power on Target (kW) 1440 880 880

Linac Beam Duty Factor (%) 6.0 4.0 4.0

Beam intensity on Target (protons per pulse) 1.5 x 1014 1.3 x 1014 1.0 x 1014

SCL Cavities in Service 81 80 80

Major Parameters achieved vs. designed


SCL status SCL is providing a very reliable operation for neutron

production following SNS power ramp-up Down time < 5 min/day (<1 trip/day)

930 MeV + 10 MeV (energy reserve)

In-situ plasma processing Initial attempt showed very promising results R&D plan for 1 year

0

2

4

6

8

10

12

14

16

18

1a 2b 3c 5a 6b 7c 9a 10b 11c 12d 13d 14d 15d 16d 17d 18d 19d 20d 21d 22d 23d

Cavity number

Ea

cc

(M

V/m

)


At the Design Phase


SNS Beam Time-Structure

Macro-pulses

Midi-pulses

Micro-pulsesTb=2.4845 ns (1/402.5MHz)

N=260 micropulses

Tib=645 ns Tg=300 ns

Ti=945 ns (1/1.059 MHz)

M=1,060 mid-pulses

Tmb=1msTG=15.7ms

Tm=16.7ms (1/60Hz)

6e8 particle/micropulse95 pC/micropulse

38mA

26mA

1.56 mA


HOM analysis for SNS

Cavity Design Finding HOM’s for the reference

geometry (R/Q, f…)

Linac beam dynamics

Ring Injection;foil, beam loss, &

activation

Finding HOM’sincluding the effects

of manufacturing imperfection

BBU for each HOM’s assuming Q

Compare resultsallowable ?

Q threshold

Qex estimation & design basis for HOM coupler

Pre-estimation of heat load as a function of Qext & Qo

YesNoReduce

Q

f measurement of non-SNS cavity Q threshold


HOM findings (I)

Monopoles, dipoles, quadrupoles, sextupoles Up to cutoff frequencies of pipes (~3 GHz) Single cavity, Superstructure r/Q’s of all modes as a function of particle velocity

Mechanical imperfection Random geometrical error based on manufacturing

experiences Possibility of having trapped modes


HOM findings (II)

0

0.2

0.4

0.6

0.8

1

1.2

-2 -1 0 1 2 3 4 5 6 7

500 mm beam pipe +400 mm conical ends 500 mm beam pipe +

400 mm conical endsStainless Steel Bellows (91 mm)

; FPC sides

0

0.2

0.4

0.6

0.8

1

1.2

-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8


E_2 0 mm

9.00E-01

9.20E-01

9.40E-01

9.60E-01

9.80E-01

1.00E+00

0.00E+00 5.00E-03 1.00E-02 1.50E-02 2.00E-02 2.50E-02

r (m)

n

HOM findings (medium)

1.E-04

1.E-03

1.E-02

1.E-01

1.E+00

1.E+01

1.E+02

1.E+03

0.5 1.0 1.5 2.0 2.5 3.0 3.5frequency (GHz)

Maximu

m R/Q (

Ohm)

1.E-06

1.E-03

1.E+00

1.E+03

1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0

frequency (GHz)

Maximu

m R/Q (

Ohm/cm

4)

1.80E+00

1.85E+00

1.90E+00

1.95E+00

2.00E+00

2.05E+00

0.00E+00 5.00E-03 1.00E-02 1.50E-02 2.00E-02

r (m )n

n vs. r; radial gradient of the axial electric field (Ezrn)

dipole

quadrupole

1.E-03

1.E-02

1.E-01

1.E+00

0.55 0.57 0.59 0.61 0.63 0.65 0.67 0.69

Beta

R/Q

(o

hm

)

mode 29 mode 30mode 31 mode 32mode 33 mode 34mode 35 mode 36

r/Q vs.

1

1

1

1


HOM findings (high)

1.E-04

1.E-03

1.E-02

1.E-01

1.E+00

1.E+01

0.7 0.75 0.8 0.85 0.9beta

R/Q

(O

hm

)

mode 31 mode 32

mode 33 mode 34

mode 35 mode 36

r/Q vs.

1.E-02

1.E-01

1.E+00

1.E+01

1.E+02

1.E+03

0.5 1.0 1.5 2.0 2.5 3.0 3.5Frequency (GHz)

Maxim

um R/Q

(Ohm

)

1.E-03

1.E-02

1.E-01

1.E+00

1.E+01

1.E+02

1.E+03

1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0Frequency (GHz)

Maxim

um R/Q

(Ohm

/cm2

)

1

1

1

1


Geometrical imperfection (for any possible trapped mode)

0Ls

s

0Ls

s

200

1

))/sin()/cos((i

sisi LsibLsia

0 . 2 0 . 4 0 . 6 0 . 8 1

- 2 0

- 1 0

1 0

2 0

Ls

1

-1

0 . 2 0 . 4 0 . 6 0 . 8 1

- 2 0

- 1 0

1 0

2 0

Ls

1

-1

Deviations from the reference geometry

; normal to the surfaceLs; total length along the cavity surface, ~1.9 m for high beta cavityai & bi; random coefficient

An example of the random mechanical perturbation.

0.130

0.135

0.140

0.145

0.150

0.155

0.160

0.165

0.170

0.080 0.090 0.100 0.110 0.120

z (m)

r (m

)

ref. shape ref. Geometry (MAFIA) P1 (MAFIA)

Trapped mode is unlikely up to quadrupole modes


Properties of HOM frequency (R. Sundelin’s study at Cornell) HOM frequency Centroid Error between analysis & real

ones Fractional error; (fanalysis-freal,avg)/fanalysis< 0.0038 SNS used 0.8 % (for conservative analysis)

put frequency centroid on the highest spectral line in the range

HOM frequency spread

fo; fundamental frequency, fn; HOM frequency For beam dynamics 20 % of this value were used for

conservative analysis

Non-pi fundamental mode

000109.0 ff n

027.0mod

mod

calculatede

e

calculated

calculatedmeasured

ff

f

f

ff

, 0.0675 was used for analysis


Bunch Tracking (I)JLab; R. Sundelin, L. Merminga, G. Krafft, B. Yunn, J. Delayan SNS; D. Jeon, J. Wei, M. Doleans, S. Kim Transverse

Cumulative effects True instability; can occur at almost any frequency

No transverse instabilities for Q<108 (20% of for f spread) Error magnification; worst when an HOM frequency

differs by of the order of 1 cavity bandwidth from a beam spectral lines Error magnification~10.00062 (0.8 % centroid error)

Longitudinal Instability

Bunch energy error, bunch-to-bunch variation Can occur at almost any frequency Non-pi fundamental passband can excite oscillations

No instability & significant of energy error at Q<108


Bunch Tracking (II) Bunch output energy errorWith artificially high r/Q for 5/6 mode

All quads are turned off for benchmarkingBeam injection; 1 mm off axis

HOM centroids; switched to the most dangerous placeBeam current; 50 mA, 1 mm off-axis injectionHOM’s Q=108


HOM induced power

Individual cavity issue

Function of HOM f, r/Q, Q

Same logic is applied for centroid

Both analytic & numeric calculation

Find the possible peak HOM induced power


Beam induced HOM Power (I)

1.E-091.E-081.E-071.E-061.E-051.E-041.E-031.E-021.E-01

1.E+001.E+011.E+02

2.8075E+09 2.8115E+09 2.8155E+09 2.8195E+09 2.8235E+09 2.8275E+09f (Hz)

402.5 MHz micro-bunch resonance1.059 MHz (1/945 ns) midi-pulseresonance

1.55 MHz (1/645 ns) midi-pulsewithout gap anti-resonance

Qex=106

f (Hz)

Tim

e A

vera

ged

HO

M

Po

wer

/(r/

Q),

[W

/Oh

m)

100

1

0.01

In continuous frequency domain

1.E-02

1.E-01

1.E+00

1.E+01

1.E+02

1.E+03

1.E+04

1.E+05

1.E+01 1.E+04 1.E+07 1.E+10

Qex

Tim

e A

ve

ra

ge

d H

OM

Po

we

r/(

R/Q

) a

t 2

.81

75

GH

z

(m

icro

-p

uls

e r

es

on

an

ce

),

[W

/Oh

m]

Micro-pulse

Midi-pulse

Macro-pulseexQQr

tVtP

2)()(

Tt

tavg dttPT

P1

1)(

1

HOM Power

Time AveragedHOM Power


Beam induced HOM Power (II)

1.E-01

1.E+00

1.E+01

1.E+02

1.E+03

1.E+04

2.817499E+09 2.817500E+09 2.817501E+09

f (Hz)

Tim

e A

vera

ged

HO

M P

ow

er/(R

/Q),

[W/O

hm

]


Dangerous Modes at around main spectral lines within the centroid error range (Medium beta cavity)

0

0.2

0.4

0.6

0.8

1

1.2

-0.6 -0.4 -0.2 0 0.2 0.4 0.6

0

0.1

0.2

0.3

0.4

0.5

0.6

-3 -2 -1 0 1 2 3

0

0.2

0.4

0.6

0.8

1

1.2

-3 -2 -1 0 1 2 3

1.E-03

1.E-02

1.E-01

1.E+00

0.55 0.57 0.59 0.61 0.63 0.65 0.67 0.69

Beta

R/Q

(o

hm

)

mode 29 mode 30mode 31 mode 32mode 33 mode 34mode 35 mode 36

Mode f (Hz) Qo (due to SS Bellows)

31-1 2.800995E+09 1.5660E+07

31-2 2.800834E+09 1.6235E+07

31-3 2.800863E+09 1.5158E+07

32-1 2.820670E+09 1.9979E+07

32-2 2.820575E+09 1.8349E+07

32-3 2.820466E+09 1.9049E+07

36-1 3.230296E+09 3.5576E+04


Dangerous Modes at around main spectral lines within the Centroid Error range (High beta cavity)

0

0.2

0.4

0.6

0.8

1

1.2

-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8

0

0.2

0.4

0.6

0.8

1

1.2

-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.81.E-04

1.E-03

1.E-02

1.E-01

1.E+00

0.7 0.75 0.8 0.85 0.9beta

R/Q

(O

hm

)

mode 25 mode 26

mode 27 mode 28

mode 29 mode 30

1.E-04

1.E-03

1.E-02

1.E-01

1.E+00

1.E+01

0.7 0.75 0.8 0.85 0.9beta

R/Q

(O

hm

)

mode 31 mode 32

mode 33 mode 34

mode 35 mode 36

Modes f (Hz) Qo (due to SS Bellows)

25-1 2.41389E+09 < 2.0E+6 (FPC)

25-2 2.41406E+09 < 2.0E+6 (FPC)

25-3 2.41394E+09 < 2.0E+6 (FPC)

25-4 2.41327E+09 < 2.0E+6 (FPC)

35-1 2.81540E+09 7.942E+06

35-2 2.81576E+09 4.206E+06

35-3 2.81555E+09 4.314E+06

35-4 2.81556E+09 2.772E+06

36-1 2.83064E+09 5.508E+06

36-2 2.83046E+09 2.655E+06

36-3 2.83016E+09 2.086E+06

36-4 2.83008E+09 2.224E+06


Maximum HOM power of each mode in the Centroid Error range for Maximum r/Q

1.E-08

1.E-07

1.E-06

1.E-05

1.E-04

1.E-03

1.E-02

1.E-01

1.E+00

1.E+01

1.E+02

5 10 15 20 25 30 35 40

mode number

Tim

e A

ve

rag

ed

HO

M P

ow

er

pe

r

Ca

vit

y f

or

the

ma

x R

/Q [

W]

Qex=10^3 Qex=10^4Qex=10^5 Qex=10^6Qex=10^7 Qex=10^8

1.E-07

1.E-06

1.E-05

1.E-04

1.E-03

1.E-02

1.E-01

1.E+00

1.E+01

1.E+02

1.E+03

5 10 15 20 25 30 35 40

mode number

Tim

e A

vera

ged

HO

M P

ow

er p

er

Cav

ity fo

r th

e m

ax R

/Q [W

]

Qex=10 3̂ Qex=10 4̂Qex=10 5̂ Qex=10 6̂Qex=10 7̂ Qex=10 8̂

Beam pipe mode;~ out of concern

FPC coupling;Less concern

Damping requirement; Medium beta cavity 10^6 High beta cavity 10^5


The decision for SNS HOM

No Beam dynamics issue

Centroid error, f spread & location of cavities; in question

When Q>105, 106, there’s a concern. HOM power ~ fundamental power dissipation but the probability is very low even under the conservative

assumptions

Extra insurance SNS is the first pulsed proton SC linac Any issues were treated in a very conservative way

Ex. Piezo tuner; we’ve never used them


SNS HOM Coupler


SNS HOM Coupler

Coaxial type notch filter scaled from TTF was chosen

end-cell

HOM can

HOM antenna

HOMFeedthrough

gap

-100

-90

-80

-70

-60

-50

-40

-30

-20

-10

0

300 800 1300 1800 2300 2800 3300 3800

Frequency [MHz]

S12

[d

B]


HOM coupler development

Identification of all HOM modes; very good agreement

Low power tests confirmed its functionality (JLab) Damping; dangerous modes to have Q<~10^5

0

0.2

0.4

0.6

0.8

1

1.2

-0.6 -0.4 -0.2 0 0.2 0.4 0.6

z [m]

E2

-0.3

0.7

1.7

2.7

3.7

4.7

5.7

0.0 5.0 10.0 15.0 20.0 25.0 30.0

Time [s]

Ph

ase

[de

g]

Without FPC f=2812.366 MHz With FPC (5mm penetration) f=2812.131 MHz

FPC side

calculation


Fundamental mode thru HOM coupler

Normal waveform of fundamental mode from HOM ports (y-axis; log scale)

HOMA

HOMB

Fundamental mode couplingHigh 1010~ 1012

< a few W during pulse


Abnormal HOM coupler signals (RF only, no beam)

1~5 Hz 10 Hz 30 Hz

~’0’ coupling and rep. rate dependent signals

Electron activities (MP & discharge; observations under close attention)


Problems (II)


Problems while running RF onlyAny electron activity (multipacting, burst of field emitter, etc)Destroy standing wave pattern (or notching characteristics)Large fundamental power couplingFeedthrough/transmission line damage (most of attenuators were blown up)Irreversible

Electric Field

Magnetic field

Gonin, N. Solyak


Turn-on and High power commissioning First turn on must be closely watched and controlled

(possible irreversible damage) Initial (the first) powering-up, pushing limits, increasing rep.

rate (extreme care, close attention) Aggressive MP, burst of FE possibly damage weak

components Similar situation after thermal cycle (and after long shut down

too) behavior of the same cavity can be considerably different from run to run

Subsequent turn-ons (after long shut-down) also need close attention: behavior of the same cavity can be considerably different from run to run gas re-distribution

Cryomodules/strings must be removed and rebuilt if vented/damaged


Large fundamental mode coupling 11b; repaired at JLab but non-operable from the beginning, no

notch 19b; March 06 turned off (10 W coupling at 1 MW/m) 3 cavities; operable but limited by HOM power

Not related with damage, just worse location of notching freq.

6 cavities; abnormal waveforms about ‘0’ coupling Seems to be a (partial) disconnection in feedthrough/cable in CM May have leak

beam line vacuum leak CM12

Largest field emission Feedthrough damaged?

Limited by Fundamental mode in HOM coupler


Re-evaluation


Re-evaluation of necessities for HOM couplers

Same arguments on BBU and instability Stainless steel bellows and fundamental power

coupler Q < 108

What is the possibility that one can have high HOM induced power? HOM verifications

HOM statistics

Beam induced signal


HOM frequency measurementat 4.4 K (between HOMA & HOMB)

Frequency measurement of the dangerous modes for all cavities in the linac tunnel

　 Medium beta cavity High beta cavityassumption for analysisMode 31 32 25 35 36

Average (GHz) 2.81197 2.83017 2.41587 2.81383 2.83018 　

Sigma (MHz) 3.19626 3.33151 1.54211 2.09127 2.71721 　

Analysis (GHz) 2.80100 2.82050 2.41400 2.81550 2.83030 　

sigma/(fn-fo) 0.00159 0.00165 0.00096 0.00104 0.00134 0.00022

fractional centroid error -0.00392 -0.00343 -0.00077 0.00059 0.00004 0.008

In addition, to have an estimation of HOM frequency shift at 2K, tests have been done for some high beta cavities by changing tuner position (~60 kHz)Frequency changes less than ~400 kHz

Out of concern2.8175 GHz

Out of concern2.8175 GHz


Zoom-in

1.E-06

1.E-05

1.E-04

1.E-03

1.E-02

1.E-01

1.E+00

1.E+01

1.E+02

1.E+03

1.E+04

2.815 2.8155 2.816 2.8165 2.817 2.8175 2.818 2.8185 2.819 2.8195 2.82f (GHz)

Pow

er/

(r/Q

) (W

/Ohm

)

Peak valueat the end of pulse

Time averaged

The Q’s of modes are less than about ~107 due to the damping on the SS bellows, there’s no macro-pulse resonance.Peak value is more meaningful.

In following pages frequency distributions are plotted on the peak power spectral lines.Beam current (38 mA peak) in the following examples


1.E-06

1.E-05

1.E-04

1.E-03

1.E-02

1.E-01

1.E+00

1.E+01

1.E+02

2.8150 2.8155 2.8160 2.8165 2.8170 2.8175 2.8180 2.8185 2.8190 2.8195 2.8200f (GHz)

Peak

Pow

er

(W)

Zoom-in (Mode 31 Medium beta)

Peak power at maximum r/Q (~0.3 Ohm)

Only 5 cavities have mode 31 in +/-2.5 MHz from main spectral line (2.8175 GHz).(Mode 25 of other 28 cavities are below 2.815 GHz) out of concern.Q’s <~10^7 (with HOM couplers 10^4)Only main spectral line will be a concern.Presently all HOM’s are far away from the dangerous beam spectral line.

Red dots just show the frequency locations

10^7

10^6

10^5


Zoom-in (Mode 25 High beta)

1.E-06

1.E-05

1.E-04

1.E-03

1.E-02

1.E-01

1.E+00

1.E+01

1.E+02

1.E+03

2.4110 2.4120 2.4130 2.4140 2.4150 2.4160 2.4170 2.4180 2.4190f (GHz)

Peak

Pow

er

(W)

Peak power at maximum r/Q (~0.35 Ohm)

All Mode 25’s are within +/-3.5 MHz from main spectral line (2.415 GHz).FPC coupling at most 10^5 (with HOM coupler 10^3)The main spectral line and the neighboring ones will be a concern Presently all HOM’s are far away from the dangerous beam spectral line.


10^6

10^5


1.E-06

1.E-05

1.E-04

1.E-03

1.E-02

1.E-01

1.E+00

1.E+01

1.E+02

1.E+03

2.8150 2.8155 2.8160 2.8165 2.8170 2.8175 2.8180 2.8185 2.8190 2.8195 2.8200f (GHz)

Peak

Pow

er

(W)

Zoom-in (Mode 35 High beta)Peak power at maximum r/Q (~0.4 Ohm)


10^6

10^5

10^7

16 cavities have mode 35 in +/-2.5 MHz from main spectral line (2.8175 GHz).Q’s <10^7 (with HOM coupler 10^4)Only first midi-pulse sub spectral line will be a concern.Presently all HOM’s are far away from the dangerous beam spectral line.


1.E-06

1.E-05

1.E-04

1.E-03

1.E-02

1.E-01

1.E+00

1.E+01

1.E+02

1.E+03

2.81620 2.81625 2.81630 2.81635 2.81640 2.81645 2.81650 2.81655 2.81660f (GHz)

Peak

Pow

er

(W)

Another concerns

10^6

10^5

10^7

More zoom-in (Mode 35 High beta)

1.E-101.E-091.E-081.E-071.E-061.E-051.E-041.E-031.E-021.E-011.E+001.E+011.E+02

2.8160E+09 2.8165E+09 2.8170E+09 2.8175E+09 2.8180E+09 2.8185E+09 2.8190E+09f (Hz)

Tim

e-a

vera

ge

d H

OM

po

we

r/(R

/Q),

[W

/Oh

m]

800 MeV

1000 GeV

1300 MeV

Q~10^6~80 kHz (Beam current)^2


w/ Beam (no measurable HOM signals from beam was observed)

Direct wakefield, just showing beam time-structure including frequencies higher than the cut-off frequency


Decision in 2007

Take out feedthrough as needed


19b TDR measurement & comparison; almost shorted Trace of discharge Dimension looks OK but large coupling HOM feed through removed Very aggressive electron activity at the HOM coupler Recovered; back in the tunnel in Feb. 08

CM12; beam line vacuum leak 12a and 12d had leaks at the feedthroughs (four out of eight)

HOM feedthroughs are removed and back to service in Feb. 09

Repair (so far 2 CMs were taken out from the tunnel)


SNS beam


Bunch fluctuation Best conservative guessing; ~0.3 % (1% will be

enough for analysis)


SNS beam time-structure


Chopper imbalance

Every 4th midi pulse; rising is slower


40-50 dB less than main line; negligible


Beam Experiences

Beam loss Much Less than 1 W/m in average Not a show stopper

Energy jitter ~1 MeV (from other sources)

No dependencies on beam loss Beam current Pulse length Pulse repetition rate As tuning is getting better, less loss/C


Summary

SNS HOM concerns & history Beam breakup & instability; no issue HOM power is the main reason for SNS to have HOM coupler

Availability & Reliability; Most Important Issue HOM couplers in SNS have been showing deterioration/failure as reported Reliability & availability of SNS SRF cavities will be much higher w/o HOM

coupler

More realistic analysis with actual frequency distributions measured. Probabilities for hitting dangerous beam spectral lines are much less.

Future concerns HOM feedthroughs will be taken out as needed PUP cryomodule

At least will not have HOM feedthroughs for cavities we already have Will not have HOM couplers for new cavities

Documents

HOM Experiences at the SNS SCL