Managed by UT-Battelle for the Department of Energy Sept. 27, 2009 Oak Ridge, TN, USA SNS External Antenna Source Problems and Solutions R.F. Welton, J.

Managed by UT-Battellefor the Department of Energy

Sept. 27, 2009Oak Ridge, TN,

USA

SNS External Antenna Source Problems and Solutions

R.F. Welton, J. Carmichael, N. J. Desai, R. Fuga, R.H. Goulding, B. Han, Y. Kang, S.W. Lee, S.N. Murray, T. Pennisi , K. G. Potter, M. Santana, and M.P. Stockli

1. Detailed source description2. Brief history of testing the AlN

source3. Problems encountered during 1st

operation of AlN source on SNS:1. Water leaks2. Antenna failures3. Magnet heating4. Plasma ignition 5. Beam decay

4. Proposed solutions

2 Managed by UT-Battellefor the Department of Energy

SNS External Antenna Source

3 Managed by UT-Battellefor the Department of Energy Presentation_name


A brief history of testing of the AlN external antenna source

· AlN Source first tested on test stand January 2008

· Early testing on stand (Jan-July 2008) focused on the elemental Cs system – 100mA achieved from Ni collar!

· Later testing on stand (July 2008 to present) focused on using baseline chromate Cs system

· Early tests on SNS front end spanned only a few days and showed that ~40mA could be transported through the RFQ (July 2008)

· 4 production versions of the source were prepared and testing on the stand began in Nov 2008

· Front end operation spanned Feb – April 2009. We noted a 97% availability and · 1 water leaks in cooling jacket (later found bolts not tight)· 2 antenna failures· 2 p-gun failures· Unexplained ~10% / week beam decay


For Ref: Operational history of the AlN source

Performance: Antenna Build gap elog entry notes beam current Current RF power edump gas steering collar f1 f2

number (mm) set/actual flow delta temp (kV) (kV) end of run notes(bbb meter) ( C )

ext3-01-14-09 0.12 cesiated on 1/27 routine source removal no beam extracted

ext1-12-11-08 0.76 2-12-09-16:08 after cesiation 40mA 540A 37/kW 6.5kV 32.8 300 ~30 46 4635mA 520A 34/ kW 6.5kV 32.8 300 ~30 46 46 routine source removal

2-13-09-12:28 32mA 520A 34/ kW 6.5kV 31.8 300 ~30 46 44 ceramic edump insulators2-13-09-17:44 recesiated 35mA 520A 35/27kW 7.0kV 27.3 500 ~30 46 39 found arc-through due to high2-16-09-06:30 edump insulators 38mA 520A 35/27kW 7.0kV 37.3 500 ~30 46 39 edump voltage

found damaged

ext2-02-12-09 0.43 2-20-09-20:10 brief tune after Cs 33mA 520A 35/27kW 5.7kV 36.2 600 ~30 48 442-23-09-12:23 2min cesiation 10-30mA2-23-09-15:43 power study: 32mA 520A 35/27kW 6.3 31.2 500 ~30 48 46 unable to maintain lite plasma

35mA 540A 40/30 kW source replaced due to38mA 560A 49/ kW antenna / pgun problems

New LEBT feedthrough issues LBNL source tested35mA / 31kW (44kW set)

ext3-02-03-09 0.36 pgun, RF amp & matching network water leak issues

2-27-09-22:58 before cesiation 35mA 520A 35/27 kW 5 45.4 200 ~30 45 44 routine source replacement:3-1-2009-13:27 recesiated 33mA 520A 35/27kW 6.2 43.8 400 ~30 45 41 high gas flow required

34mA 600A 58/ kW 6.2 43.8 400 ~30 45 41 backflange Ta shield not installed

ext1-02-20-09 0.76 3-3-09-15:09 alignment brought +7mA 35mA 540A 38/26kW 6.2kV 27.3 200 ~40 48 45 routnie soruce replacementno findings upon source inspection

ext2-02-26-09 0.6 3/9/2009 32-33mA 600A 46/30kW 6.2 42.7 300 40 47 463-11-09-12:38 after cesiation 35mA 560A 39.5/kW 6.2 41.6 100 48 47 46 routine source replacement 3-16-09-17:50 after cesiation 35mA 480A 32.5/kW 6.2 38.3 0 44 48 46 high H2 flow - outlet gap found

ext1-03-11-09 0.46 3-16-09:17:50 source conditioned w 560A 35mA 480A 32.5/kW 6.2 38 0 44 48 46 split backflange ring collapsed intoantenna resulting in antenna damage

pgun did not lite

ext2-03-18-09 0.61 3-25-09-10:59 source turn-over 36mA 540A 39/kW 6.2 27 100 47 48 46 routine source change3-26-09-10:35 1st heating of collar 37mA 580A 41.5/kW 6.2 24 200 150 48 46

ext1-03-26-09 0.66 4-6-09-19:24 source ready for ops 36mA 520A 37/kW antenna failure

ext2 -04-07-09 did not run water leak in Lexan (bolts not tight)

ext3 - 03-06-09 4-8-09-1:12 source ready for ops 36mA 580A 43/kW 6.2 32.3 300 155 48 46 antenna failure4-8-09-8:45 38mA 500A 6.2 32.8 300 149 48 46

ext2-04-07-09 0.61 4-13-09-19:36 source ready for ops 35mA 520A 39/kW 6.2 33.9 500 155 48 46 pgun failure

ext1-04-09-09 0.56 4-19-09-21:03 two-layer antenna 35mA 540A 31.5/kW 6.2 28.7 0 145 48 464-20-09-10:03 edump resistor failure routine source change


Improved Plasma Chamber Cooling Jacket

• Early versions were made from unannealed polycarbonate which developed small fractures from residual stress and deformation under water pressure leading to small leaks which would terminate source operation

• The current version employs a both dimensionally stable PEEK and stainless steel materials, see photo below

• Jacket is performing well during initial testing on the test stand

• No failures of revised water jackets yet noted

Deformation

Stress

Original (Blue)

Revised (Red)


PEEK Cooling Jacket and SS Manifold


Improved RF Antenna and Ferrite Backing

• Several antenna failures were noted during operation - burning of the polyolefin insulation

• It was discovered that during antenna fabrication inner and outer turns were mixed resulting in RF electric fields which exceeded the breakdown strength of air.

• Inner and outer turns were then separated (as in the original design) using a 3/32 inch Teflon separator and no failures were noted thereafter over several experimental runs

• To further reduce the peak and overall electric fields of the antenna a 3.5-turn antenna and Teflon holder have been designed

• Alternatively, potting the antenna also remains a promising approach

• Ferrites backing the antenna will enhance the inductive coupling efficiency with the plasma and are in new design


Improved Magnet Cooling and Confinement• Multicusp confinement magnets in the external antenna source

occasionally overheat. New magnet configurations provide better cooling as well as higher magnetic confinement fields. Combines magnet holders and water manifold.


Improved Magnet CoolingOctapole Hexapole• RF power

dissipation on the surfaces of the magnet holder was calculated from CST MICROWAVE STUDIO and used as loads for Cosmos Floworks….

• Thermal simulations show that for a net RF power of 50 kW and cooling of 1.5 GPM the maximum temperature where the magnets contact the holder as <50 ºC and ΔT of water as 1.5 ºC.

• Actual temperature is less since the plasma and ferrites will significantly reduce heating.

Toward ANT138 W

Circumferential (outer) 111 W

Total Surface Power460 W

Circumferential (inner)121 W


Improved Magnet Confinement

Current Hexapole Proposed Octapole Proposed Hexapole

• The new configuration offers 1.7 (octapole) and 2.3 (hexapole) -fold increase in magnetic plasma confinement fields as well as improved cooling.


Improving Plasma Ignition…

• Reliable plasma ignition remains as a major problem with this source• Currently CW plasma guns (shown above) affixed to the rear of the

source• Alternatively CW 13 MHz can be used for plasma ignition.

13 Managed by UT-Battellefor the Department of Energy Presentation_name

Improving Plasma Ignition

• Plasma guns based on Cu cathodes have shown long-lifetimes but are not ideal since they sputter-inject noticeable layers of Cu into the plasma chamber. Figure shows gun emission current for a 4-month run.

• Mo cathodes offer less sputtering but some seem to exhibite an unexplained poisoning effect after 1-2 weeks of operation rendering the source inoperable.

0

0.1

0.2

0.3

0.4

0.5

0.6

3/28 4/17 5/7 5/27 6/16 7/6 7/26 8/15

Date

Co

lle

cte

d c

urr

en

t

(mA

) • The application of 13 MHz RF directly to the primary antenna requires high RF CW power-levels of 0.5-1 kW (increasing with time) for reliable ignition and suffers from cross talk issues between the 2 and 13 MHz systems.


Several approaches are being pursued: • Identifying and testing, on a multiport chamber

(shown to the right), long-lived, low-sputtering cathode materials for the existing plasma guns, which includes W, Ni and SiC based on feedback from the thyratron community.

• Obtaining a commercial electron gun (see Heatwave slides)

• Operating the pulsed 2 MHz system also in a low-level, cw mode.

• An 13MHz RF plasma gun has been designed to replace the existing DC plasma gun on the source.

Improving Plasma Ignition


Improving Plasma Ignition • An RF plasma gun will take advantage of the existing 13MHz RF system used with

the baseline source. This should allow significant decoupling of the 2 and 13 MHz RF systems, reduced RF power requirement due to operation closer to the Paschen minimum (higher pressure) and power deposition into a separate cooling system.

RF Plasma Gun

0

200

400

600

800

1000

1200

1400

1600

0 2 4 6 8 10

Pressure (Torr)

Bre

akd

ow

n F

ield

Str

eng

th (

V/c

m)

Plasma chamber of source


Electrons emitted from the directly-heated dispenser cathode (red), are focused by the low-albedo, water-cooled, back-plane repeller (blue), so that they flow to the right through the non-intercepting aperture (green) into the main ionization chamber.

Ions flowing through the aperture to the left also are focused so that they miss the cathode and impinge on the back-plane repeller.

Secondary electrons produced by ions striking the back-plane repeller add to the constant current supplied by the directly-heated cathode. Since the low-temperature, long-life cathode produces a constant supply of electrons, any variation in this secondary electron emission current is inconsequential.

CONFIDENTIAL AND PROPRIETARY

Figure 1. 2-inch Diameter x 3-inch Long Optics Model

Cathode current is controlled by the potentials, jCATH and jBPR

Electron and ion focusing is controlled by the potentials, jCATH and jBPR

These potentials can be established using a single power supply and voltage divider

Improving Plasma Ignition – Commercial Options – AS&E, Heatwave Proposal


• Long life, reproducible, reliable directly-heated dispenser cathodes1

– 900-1200°C operation– 3 A/cm2 or more CW. > 40,000 hours lifetime– Stanford Linear Accelerator (SLAC) has 8 A/cm2 for over 45,000 hours– Reproducibility is well proven– Can be let up to air and reactivated repeatedly

• The actual design will be scaled to fit the existing system and optimized by determining, in axis-symmetric space, the actual plasma emission surface meniscus formed at the non-intercepting aperture by balancing:– The ion current density available from the plasma

• Bohm relationship

– The space-charge limited ion extraction current density

• Child’s law

CONFIDENTIAL AND PROPRIETARY

Improving Plasma Ignition – Commercial Options – AS&E, Heatwave Proposal


Integrated Solid Model of Improvements

RF Plasma Gun

New external antenna plasma chamber assembly


Summary · We have focused on addressing each issue which arose during FE operation · An integrated model of the complete core assembly has been developed· Currently changes are being reviewed and detailed by the SNS Mech Group· Looking forward to collaborating with this community!


Remaining issue: Beam Decay with time (~10% / week)

· Cause unknown

· Could result from changing Cs conditions– Impurities

· From plasma gun 5 W (Cu Mo) ?· Leaks?· Out gassing of something hot e.g an o-ring?· AlN tube?· Gasses from hot LEBT?

– Cs migration to/from AlN tube?

· Could result from changing plasma conditions– Changing coupling?

· Amplifier / matching network / antenna

· Need to look carefully at the data

Managed by UT-Battelle for the Department of Energy Sept. 27, 2009 Oak Ridge, TN, USA SNS External Antenna Source Problems and Solutions R.F. Welton, J.

Documents

antenna failures

aln source slide

department of energy

source operation

operation of aln source

department of energy

red slide

new design slide