Status of HOMS Spectra Measurements in 1.3 GHz Cavities

Status of HOMS Spectra Measurements in 1.3 GHz Cavities for LCLS-II

Andrei Lunin(on behalf of TD/SRF Group, Fermilab)HOMSC2018 ICFA Mini Workshop1-3 October 2018

FERMILAB-SLIDES-18-121-TD

This document was prepared by [COLLABORATION NAME] using the resources of the Fermi National Accelerator Laboratory (Fermilab), a U.S. Department of Energy, Office of Science, HEP User Facility. Fermilab is managed by Fermi Research Alliance, LLC (FRA), acting under Contract No. DE-AC02-07CH11359.

This document was prepared by [COLLABORATION NAME] using the resources of the Fermi National Accelerator Laboratory (Fermilab), a U.S. Department of Energy, Office of Science, HEP User Facility. Fermilab is managed by Fermi Research Alliance, LLC (FRA), acting under Contract No. DE-AC02-07CH11359.

--------•

@ENERGY Office of Science

~ ~

~· & ,. NATIONAL -- -- e ACCELERATOR ~ ... ~~ LABORATORY

CFermilab

•

10/01/2018

Outline

HOMSC2018 ICFA Mini Workshop

• Overview of the LCLS-II 1.3 GHz Cryomodule (CM) assembly

• CM RF quality control (RF/QC) procedure at Fermilab

• Measurement of HOM spectra at CM Test Facility (CMTF)

• HOMs data processing

• Operating passband statistics

• Dipole HOMs statistics

• Monopole HOMs statistics

• Cavity vendors performance

• Summary

2-----------------------0 Fermilab

Motivation

10/01/2018 HOMSC2018 ICFA Mini Workshop

HOMSC2014, A. Sukhanov et., all

???

• Measurement of actual HOMs parameters @2K insures a reliable model of resonant HOMs excitation in the LCLS-II linac

• HOMs statistics allows to estimate and to compare a consistency of cavities mechanical tolerances for different vendors

3

Motivation

• SRF cavities are very good resonance systems with multiple eigenmodes (HOMs) with very low losses (high Q-factors)

• Beam of charged particles interacts with HOMs in SRF cavities

► Single bunch interaction

- incoherent losses and wake fields

► CW beam may have beam harmonics closet

- resonance excitation of HOMs ~~~~~ ,,'r ,, ,, - at exact resonance beam power loss may be high ,,,, ,,

• for monopole modes: Plass = /1~ (R/Q)r@

► For a single cavity analysis of non-propagating modes is sufficient

OFermilab

10/01/2018

LCLS-II Linac


• 4 GeV CW SRF Linac: 35 (+5) 1.3 GHz CMs , 50/50 by JLab & Fermilab

• Single 1.3 GHz CM: eight 9-cell TESLA-type cavities (280 cavities in linac)

• CW Beam: 0.1-0.3 nC bunch charge, 1 MHz rep. rate, 0.3 mA (max) average beam current

• HOM damped Q-value: ≤ 106 for HOMs with high R/Q

4

LO L1 L2 L3 (fJ = * (p =- 21°

HL (fJ = - 21° (fJ = 0

V0 =94 MV V0 =223 MV V0 =1447 MV V0 =2409 MV /pk = 12 A /pk = 12 A (p =- 165° /pk = 50 A /pk = 1.0 kA

Lb = 2.0 mm Lb =2.0 mm V0 =55 MV Lb= 0.56 mm Lb = 0.024 mm

CMOl CM2,3 3.9GHz CM04 .. CMlS CM16 .. CM35

Functional Requirements Specification Document Document Title: 1.3 GH z Supe,rconducting IRF Cryomod ule Document !Number: LCLSll-4.5-FR~0053-R1 Page 3 of 15

1 Introduction

Fermi ab and Jefferso Lab are col laboratiing w· h SLAG on their LCLS-I111 upgrade project by

supplying SRF design and fabricaitiion expertise W- The .aocelernting strudures willl utilize

OFermilab

10/01/2018

1.3 GHz LCLS-II Cavity Coupler Ports Changes

HOMSC2018 ICFA Mini Workshop 5

HOM antenna FPC antenna

Ø11 x 0.1 mm Ø7.8 x 0.5 mm

ILC LCLS-II ILC, XFEL LCLS-II

• Smaller size of HOM antenna tip and larger the f-part gap for minimization of local RF losses

• Weaker FPC coupling (~ by one order) for matching with beam current and lowering input RF power

Changes to accommodate CW operation:

Original antenna Antenna tip cut by 8.5 mm ..,....---- -,

I O ext ~ 4e7 I

OFermilab

CM

ass

embl

y R

F/Q

C a

t Fer

mila

b

10/0

1/20

18H

OM

SC20

18 IC

FA M

ini W

orks

hop

Cavi

ty R

F/Q

C:

RT

Spe

ctra

HO

M N

otch

Ca

bles

Inte

grity

2K

Spe

ctra

FP

C tu

ne-u

p

2K H

OM

Spe

ctra

MP9

ICB

rece

ivin

gIn

com

e RF

/QC)

IB1

(VTS

) ship

ping

CMTF

IB4

6

J: ... ::, 0 V)

I TOTIJNHEL

- I. ~

[

WEST

SRFCAVITY VACUUM BAKE FURNACE AREA

EAST

CLOSED AREA

OPENAAEA

CLEANROOM DOORS

t I I I

.l'a" •' '"J

ANTE

SRFCAVJTY 120C BAKE OVENS AREA

CLEAN ROOM CLASS 1000 (311FT. X 24 FT.)

-·~ 200 FT. CLASS 1000

SOFTWW. CLEAHROOM

t l I

[ :.:

·"'7 Q)

E

~ ·.::: I 0.. ~

t: ;I (") ~ ' •Ill ~ i (/) ~ s ~ ~

w "' ----i "' w >

~ ,.

I~ 5 .

j V I :l " ' >I

a:

" @ g <>

:i:

I i i '.J.,j 09

1 JGHz.CAVITY HELIUM VESSEL WELDING AREA

"" !

25 FT.

,~E !I •~ I

CAVITY STRING ASS EM Bly iEA 5 i fs'oE~:~~.fLASS100 ~ so ~

Cl,.E#JIIIQOMMI HAIIDUMOUNIT

ONE 25 TON CRANE

z 0

~ J:

O> 0 ,, ,..

.c ca -■-E 1-

.f <>

Cryomodule Cold Test @CMTF


HOMs 2K Spectra Measurements

Cables: 20m ⅜” Heliax, ~3dB lossHOM1 to HOM2 S21 signalAgilent E5071C NWA, <4.5GHzLabVIEW data acquisition Frequency sub-ranges: 5Step: 0.25 .. 2 kHzMax Qext: 5E6

7-----------------------0 Fermilab

Frequency [MHz]1500 2000 2500 3000 3500 4000

S 21 [d

B]

-100

-80

-60

-40

-20

0

Rǁ/

Q, R

⊥/Q

, [W

, W/c

m2 ]

10-2

10-1

100

101

102

103S21 raw dataR/Q - Dipole Bands 1&2&3R/Q - Monopole Bands 2&3 R/Q - Monopole Bands 4&5

HOMs 2K Spectra @CMTF


• High risk HOMs: Dipole bands 1&2&3, Monopole band 2&5• Dipole #3 and Monopole #5 bands interferes with other HOMs (difficult to sort out)• Logged CM HOMs data: Dipole bands 1&2 and Monopole band 2

8

Frequency [MHz]1500 2000 2500 3000 3500 4000

QEX

T

1e+3

1e+4

1e+5

1e+6Dipole Bands 1&2&3&4Monopole Bands 2&3Monopole Band 5Quadrupole Band 1

•

• I I • • \. ' ~: -

-· E •·

• • • •

• I • ., , . ..... ~--.-

• • • •

-----------------------0 Fermilab

Operating Passband 2K Spectrum


• Operating band monopole modes may generate up to 1 W power, if in resonance with the harmonic frequencies of 0.3 mA average beam current

• High Q-value (up to 109): difficult to locate and measure• Slow S21 data acquisition: BW <20 Hz• We take data for 3 cavities in the CM (left, right and middle)

9

~ 1W

l.000E+09

<Q> 1.000E+08 I l.000E+07 I I ■ ■ • • ■

1 2 3 4 5 6 7 8 9

1.00E+07 R/Q<Q> ---------- --- ' / ' { l

\ /

1.00E+06 ' ..... --- -----

1.00E+0S

1.00E+04

1 2 3 4 5 6 7 8 9

-----------------------0 Fermilab

HOMs Data Processing

10/01/2018 HOMSC2018 ICFA Mini Workshop 10

• Data filtering issues: noise, spurious peaks, bands overlapping• Final sorted HOMs data are attached to each CM traveler

Mathcad script for finding F&Q

Sorted HOMs data

10~-~ ~ ~ ~ - ~~---- ----- - -----~ ~ ------------------+-----+~ ---i

9--

8

7

6

' 5

3

2

O'------'

1e+9 ---------------------------

j 1e+8

1e+7

.. • '

1275

• ~ • . \ -

• CAVITY#1 . CAVITY#2 . CAVITY#3 . CAVITY#4 . CAVITY#S . CAVITYl6 . CAVITY#7 . CAVITY#8

. l e+6 • . . . . . .

CAVITY#1 CAVITY#2 CAVITYt3 CAVITYU CAVITY#S CAVITYl6 CAVITYt7 CAVITY#S

--~---------~---------~------ 1·

' • le+S -- ',:.-L ---~--------- ,·---------+---------

LCLS-11 Cryomodule Testing at CMTF (RFCG) 464255 Rev. G

.; . , • . I . ..

' . • i . • •

le+3

1280 1285 1290 1295 1300 1600 1650 1700 1750 1800 1850 1900

Freq., [MHz] Freq. , [MHz]

e+6

r e+5

'

i ' ' ----+-----t-----i-----r--

! I ! . . I

i,--,.,-....,.,,.-----,- 20

. CAVITYl1 . CAVITY#2 . CAVITY fl . CAVITYIM . CAVITYU . CAVITY IN . CAVITYf7 . CAVITY ff

'

- 30

- 40

A Et

- 50

- 60

- 70 :!. 9" 103

' ! I ' ------t----' ' I ' I I • ... t-----➔-----~ ~----.. .. ...

.. --- . ' . '

------1,-----L, ____ J __ • ___ •L, ---e .J, _____ :-• e+4 . - . -~-----!-----.J-----!-----

2360 2370 2380 2390 2400 2410 2420 2430 2440 2450 2460 2470

Freq., [MHz]

OFermilab

Statistics: Operating Passband Frequencies


• Cavity Vendors: pCM01 – 8 by AES, 9 production CMs – 50 by RI & 22 by EZ• AES cavities were procured for the ILC prototyping and passed many clean&tune steps• There is a large systematic frequency span between RI & EZ spectra (dies error, e/b welding ?)• The average length of RI’s cavities is ~1.5 mm shorter than the EZ’s after field flatness tuning

Cavity Length Error [mm]

-2 -1 0 1 2

Cou

nt

0

2

4

6

8

RIEZ

RI, EZ - Monopole Band 1

1.400 ■ EZ 1.200

~ 1.000 ■ RI

~ 0.800 ■ <EZ>-<RI> Z 0.600

II ~

II II D VI 0.400

I l l •• 1 0.200 II --· 0.000 --1 2 3 4 5 6 7 8 9

Mode#

RI, EZ, AES Monopole Band 1

1.000 ■ EZ

~ 0.800 ■ RI N

■ AES :::c ~ 0.600

G: Q.400 0 I-VI 0.200

I■ I■ I■ 0.000 •• ·- -- - - -1 2 3 4 5 6 7 8 9

Mode#

OFermilab

Statistics: Operating Passband Q-loaded


• Both RI and EZ spectra deviate from the spectrum of ideal cavity used in HFSS• There is a good agreement of average Q values with HFSS simulations

# 1 2 3 4 5 6 7 8 9F [MHz] 1276.62 1278.67 1281.82 1285.73 1289.91 1293.87 1297.12 1299.25 1300.00

Q-loaded 6.71E+8 1.73E+8 8.13E+7 4.92E+7 3.45E+7 2.69E+7 2.27E+7 2.07E+7 4.0e+7

ANSYSHFSS

<Q>

1.000E+08

I 1.000E+07 I I I I I I I 1274.656 1276.883 1280.313 1284.981 1288.584 1293.337 1296.870 1299.198 1300.000

■ EZ 5.402E+08 1.464E+08 7.445E+07 4.941E+07 3.407E+07 2.797E+07 2.478E+07 2.418E+07 4.000E+07

■ EZ 7.058E+o8 2.014E+o8 9.004E+o7 6.435E+o7 4.038E+o7 3.495E+o7 3.000E+-07 3.021E+o7 4.000E+-07 QMAX

<Q>

1.000E+08

I 1.000E+07 I I I I I I I 1275.974 1278.113 1281.362 1285.383 1289.659 1293.716 1297.052 1299.248 1300.000

■ RI 5.907E+08 1.510E+08 6.438E+07 4.494E+07 3.359E+07 2.785E+07 2.568E+07 2.402E+07 4.000E+07

■ RI 9.519E+o8 2.616E+o8 9.547E+o7 6.001E+o7 4.580E+o7 3.391E+o7 3.094E+o7 3.250E+o7 4.000E+-07 ~AX

OFermilab

HOMs Statistics: Monopole Passband #2 Frequencies


Frequency [MHz]

2445 2450 2455 2460 2465 2470

Cou

nt

0

2

4

6

8

10

12

14

RIEZ

Mode #9

12.000

10.000 ,. ::c: 8.000 ~ > 6.000 w C 4.000 I-VI

2.000

0.000

35.000

30.000

,. 25.000 ::c: ~ 20.000

Z 15.000 ~ VI 10.000

5.000

0.000

1. 11

I■ 11


I . 1. I. 1. 1. 12 13 14 15 16

Mode#


1.1 111 1.. 1.. 1 •• 12 13 14 15

Mode#

16

■ EZ ■ RI ■ AES

I■ I■ I■ 17 18 19

i D

t ■ EZ ■ RI ■ <EZ> - <RI>

I■■ 111 111 17 18 19

vFermilab

Q-loaded1e+4 1e+5

Cou

nt

02468

10121416

RIEZ

HOMs Statistics: Monopole Passband #2 Q-loaded


~ 10W

Mode #9

• 5 RI cavities have high Q-loaded of the M2_9 mode, EZ’s are OK

1E+6

1E+5

1E+4 I I I I <; I I I I 1E+3

2.69SE+o6 5.475E+o5 3.127E+o5 1.311E+o5 7.336E+o4 2.803E+o4 3.660E+o4 1.078E+o5 6.825E+o4 ~AX ~ .

2370.886 2377.146 2386.554 2398.397 2411.573 2425.189 2437.375 2446.661 2452.890

■ EZ 6.134E+05 1.935E+05 1.056E+05 3.694E+04 3.229E+04 1.347E+04 2.250E+04 3.624E+04 4.339E+04

<Q> 1E+5

1E+4 I I I I I I I I I 2.202E+o6 7.112E+oS 7.490E+o5 2.776E+o5 4.760E+o4 3.404E+o4 7.511E+o4 1.906E+o5 4.854E+o5

1E+3 ~AX----~--

2383.383 2388.410 2396.326 2406.715 2419.070 2432.331 2444.692 2454.672 2460.997

■ RI 4.184E+05 1.299E+05 6.128E+04 4.813E+04 1.544E+04 1.720E+04 2.967E+04 4.452E+04 1.002E+05

1E+8

1E+7

1E+6

1E+5

1E+4

1E+3

1E+2 2375 2380

I 2385

I 2390 2395 2400

R/Q*Omax

I I I 2405 2410 2415 2420 2425 2430 2435 2440

D

2445 2450 2455 2460

OFermilab

HOMs Statistics: Dipole Passband 1&2 Frequencies


The triangle shape of RI and EZ dipole spectra frequencies span might indicate the difference of the dies aperture between vendors

Frequency [MHz]

1870 1871 1872 1873 1874 1875C

ount

0

5

10

15

20

25

RIEZ

Mode #28

Sensitive to cell’s apertures

RI, EZ, AES Dipole Bands 1&2, Frequencies Deviation

6.000 ■ EZ

~ 5.ooo ■ RI

; !Iii II II II II II II II II II II II II II II II J . .t 11I .. . J 1.J 111 II 1J 11 .. t , I -~ •• •• I_,-~; .... l .. _j 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

12.0~ V V RI, EZ Dipole Bands 1&2 Frequencies Span

■ EZ

10.000 ~ ■ RI

1 II 111 ~ ~ II I I 11 11 ~I 11 I~ 111 I~ I~ Ill I~ 11 1II Ill I~ Ill Ill I~ di 1l1 1l1 11, .i: .;:z~: ~:~ ,I. L I_ 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

Mode#

D

--------------OFermilab

HO

Ms Statistics: D

ipole Passband 1&2 Q

-loaded

10/01/2018H

OM

SC2018 IC

FA Mini W

orkshop16

1e+41e+5

1e+6Count

0 5 10 15 20R

IEZ

Mode #28

RI

EZ

... ... ... ... ... ... ... ... ... m m m m m m m m m + + + + + + + + +

I-' I-' I-' I-' I-' I-' w .I>, u, a, -.J w .I>, u, a,

0 0 0 0 0 0 0 0 0 0 0 0 m m m m m m 1617.127 1605.335 + + + + + + 0 0 0 0 0 0 I-' N w .I>, V, a,

I-' 1617.679 1605.897 a, I-' w

1624.744 1613.786

I 1625.326 1614.335

I-'

I i w 1637.442 1627.828

1638.051 1628.301 I-' I I a, --.J 1655.288 1646.672 .I>,

I I 1655.956 I 1647.075 I

I-' --.J

I I I I

1677 .536 r:::::a 0 1669.837 V,

1678.229 1670.129

I-' I --.J

I

1703.212 1696.068 w V,

1703.942 1696.530

I-'

I I 1730.881 1724.406

--.J ;o a, a, ...........

0 1731.587 1724.754

□ * 1158.790 E772 I 1753.079 __ 1 I-' I 0 --.J 3 I □ 1753.453

I IJ:)

I 0 a, "' X 1759.491 '

I 0

I I 1786.015

QJ

I-' 1781.613 - 0.

00

1786.784 c::::::::z I 0 (1)

N 0. --.J

I 1782.106 -.,

I 0 QJ

1795.814 0. 1792.986 (1) I-' I 00

0. V, /--.., 1796.519 1793.368 00

I ) 1836.975 1831.388 ■ ■

I-' -,- ~ :t> gg

l I QJ <

1837.176 1831.565 (1)

m ~- ""' 3

QJ

1851.961 1848.165 O'Q

C (1)

3 1852.201 ■ ■

1848.457

s :t> 1863.928 < 1861.436 QJ (1) ~- ""' 1864.164 3

QJ 1861.763 O'Q C (1)

1872.599 3 1871.212

1872.858 1871.419

1878.654 1877.920

1878.873 1878.128

1882.705 1882.082

.,, I I I I 1882.938 1882.690

(1) I 1 1 1885.192 1884.851

"'"' 3 I I 1 1 1885.375 1885.261

=1 I 1886.411 1886.129

~I I 1886.687 1886.513

10/01/2018

Cavity Vendors Evaluation

HOMSC2018 ICFA Mini Workshop 17

VendorMonopole Operating Band

Dipole Bands 1&2

Monopole Band 2

rms span rms span rms span

RI <0.20 <0.6 <1.5 <8 <2 <9

EZ <0.15 <0.5 <1.9 <6 <5 <18

HOMs frequencies [MHz]

VendorMode 7/9Operating Band

Mode #14Dipole Bands

Mode #9Monopole Band 2

rms max rms max rms max

RI 2.5E7 3.0E7 4.6E4 6.0E5 1.0E5 4.9E5

EZ 2.4E7 3.1E7 3.8E4 7.8E4 4.3E4 6.8E4

Danger HOMs Q-loaded

-----------------------0 Fermilab

10/01/2018

Cavity Vendors Evaluation cont.


HOMs frequencies Both RI and EZ demonstrate similar HOMS frequencies r.m.s.

There are significant offsets between RI and EZ HOMs spectra, few time larger than frequencies rms for operating and dipole bands.

RI cavities are longer than EZ by about 2 mm

HOMs Q-loaded Both RI and EZ demonstrate similar Q-loaded for the operating band

Few RI cavities have significantly high Q-values of danger dipole and monopole HOMs

Maximum measured danger HOMs Q-loaded are about 2-3 times larger than ones predicted by simulation for the ideal cavity geometry.

Danger HOMs Q-values are below the LCLS-II requirements (<106)

18-----------------------0 Fermilab

10/01/2018

Discussion


Danger HOMs: 7/9 pi, dipole 2nd #14, monopole 2nd #9

Shall we try to track dipole 3rd and monopole 5th passbands? How danger they are?

HOMs spectra of RI and EZ vendors are significantly different

Is there an optimum solution how to put them in a linac? Randomly? Neighbor cavities of different vendors might reflect the HOM signal of propagating modes. Too late to change for the LCLS-II but we can think about for the ILC

Control cavity production stabilities

HOMs spectra are a sensitive indicator of the cavity imperfections. Shall we ask vendors to perform HOMs 2K measurement at VTS/HTS and to sort out/reject “bad” cavities?

Can we estimate operating mode field flatness ?

How we may use the accumulated HOMs data?

Coherent HOMs excitation might be an issue for the high-current or large scale machines (ILC). Using measured HOMs data for the beam dynamics analysis would secure our understanding of machine operation and ensure project reliability

Verification of numerical simulations results is a challenging question. Experimental HOMs data is a reliable source of the HOMs statistical parameters.

What else …?

19-----------------------0 Fermilab

10/01/2018

Conclusions


• Production of 1.3 GHz LCLS-II cryomodules is ongoing at Fermilab, 11 CMs passed the final cold test @CMTF facility

• RF/QC procedures are implemented through the CM assembly and testing

• HOMs signals acquisition procedure during the CM cold test is developed along with further data processing

• Both RI and EZ cavity vendors show similar results in terms of HOMs frequencies deviations

• Few RI cavities have a larger spread of HOMs Q-loaded

• Most of danger HOMs are identified and tracked

• Danger HOMs Q-values are below the LCLS-II specification (<106)

20-----------------------0 Fermilab

10/01/2018

Acknowledgments


We would like to thank Elvin Harms and the CMTF team for their support and assistance during the 2K HOMs spectra measurements of LCLS-II cryomodules.

21-----------------------0 Fermilab

Documents

Status of HOMS Spectra Measurements in 1.3 GHz Cavities