Characterization of plasma parameters, first beam results, and status of electroncyclotron resonance sourceS. K. Jain, Akhilesh Jain, P. R. Hannurkar, and S. Kotaiah

Citation: Review of Scientific Instruments 78, 053301 (2007); doi: 10.1063/1.2735629 View online: http://dx.doi.org/10.1063/1.2735629 View Table of Contents: http://scitation.aip.org/content/aip/journal/rsi/78/5?ver=pdfcov Published by the AIP Publishing Articles you may be interested in Wave frequency dependence of H ion production and extraction in a transformer coupled plasma H ion source atSNUa) Rev. Sci. Instrum. 83, 02A727 (2012); 10.1063/1.3678659 Beam current enhancement of microwave plasma ion source utilizing double-port rectangular cavity resonatora) Rev. Sci. Instrum. 83, 02B316 (2012); 10.1063/1.3673625 Electron cyclotron resonance plasma production by using pulse mode microwaves and dependences of ionbeam current and plasma parameters on the pulse conditiona) Rev. Sci. Instrum. 83, 02A324 (2012); 10.1063/1.3669792 Design and characterization of 2.45 GHz electron cyclotron resonance plasma source with magnetron magneticfield configuration for high flux of hyperthermal neutral beam Rev. Sci. Instrum. 81, 083301 (2010); 10.1063/1.3477998 H extraction from electron-cyclotron-resonance-driven multicusp volume source operated in pulsed mode Rev. Sci. Instrum. 77, 03A512 (2006); 10.1063/1.2165270

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:

117.240.185.82 On: Thu, 12 Jun 2014 09:57:43

Characterization of plasma parameters, first beam results, and statusof electron cyclotron resonance source

S. K. Jain,a� Akhilesh Jain, P. R. Hannurkar, and S. KotaiahAccelerator Programme, Raja Ramanna Centre for Advanced Technology, Indore 452 013, India

�Received 23 November 2006; accepted 1 April 2007; published online 2 May 2007�

Electron cyclotron resonance �ECR� plasma source at 50 keV, 30 mA proton current has beendesigned, fabricated, and assembled. Its plasma study has been done. Plasma chamber was excitedwith 350 W of microwave power at 2450 MHz, along with nitrogen and hydrogen gases.Microwave power was fed to the plasma chamber through waveguide. Plasma density and electrontemperature were studied under various operating conditions, such as magnetic field, gas pressure,and transversal distance. Langmuir probe was used for plasma characterization usingcurrent-voltage variation. The nitrogen plasma density calculated was approximately 4.5�1011 cm−3, and electron temperatures of 3–10 eV �cold� and 45–85 eV �hot� were obtained. Thetotal ion beam current of 2.5 mA was extracted, with two-electrode extraction geometry, at 15 keVbeam energy. The optimization of the source is under progress to extract 30 mA proton beam currentat 50 keV beam energy, using three-electrode extraction geometry. This source will be usedas an injector to continuous wave radio frequency quadrupole, a part of 100 MeV proton linac.The required root-mean-square normalized beam emittance is less than 0.2� mm mrad. Thisarticle presents the study of plasma parameters, first beam results, and status of ECR protonsource. © 2007 American Institute of Physics. �DOI: 10.1063/1.2735629�

I. INTRODUCTION

The electron cyclotron resonance �ECR� plasma sourceshave been widely used for producing multiple and singlecharged ion beams.1,2 The ion beams have various applica-tions for plasma processing, viz., thin film deposition,3

sputtering,4 molecular beam epitaxy,5 etching,6 ionimplanter,7 as proton beam injector to radio frequencyquardrupole,8–17 etc. The ECR plasma source at 50 keV,30 mA proton current has been designed and constructed.The source plasma parameters have been measured with aLangmuir probe. First beam results with a two-electrode ex-traction system are presented. The required value of rms nor-malized beam emittance is less than 0.2� mm mrad. Thesource will be used as an injector for 100 MeV proton linac,proposed to be constructed as initial part of a 1 GeV protonsynchrotron.18 The key components of ECR plasma sourceare plasma chamber, vacuum system, microwave system, so-lenoid magnet, plasma diagnostics devices, beam extractionelectrodes, and a beam measuring device. The plasma cham-ber was coupled to microwave circuitry through a waveguideand a quartz window. Plasma chamber was excited with350 W of microwave power at 2450 MHz, with the presenceof nitrogen and hydrogen gases. The solenoid coils are usedfor confinement of plasma and satisfy ECR resonance con-dition at the plasma chamber extremities. Detailed descrip-tion and measured results for this source are presented in thisarticle. Results are satisfactory and favorable for extractionof more beam current, specified as design specifications.

II. SOURCE DESIGN

The schematic diagram of ECR proton source with 90°mass analyzing magnet is shown in Fig. 1. The design pa-rameters of the ECR source are presented in Table I.

A. Plasma chamber and vacuum system

The plasma chamber performs two roles: it couples mi-crowave power to plasma and it contains plasma. The dimen-sions of plasma chamber are 100 mm in diameter and200 mm in length and it is used to accommodate the solenoidcoils and extraction geometry. The plasma chamber and as-sociated vacuum system were fabricated using stainless steel.The vacuum system was designed and fabricated as a “crosstype,” consisting of standard knife-edge conflated flangeCF203, CF152, and six CF70 diagnostics ports. These portsare used for connection of pump �bottom�, plasma chamber�side�, and view port �other side�, pressure gauge and feedthrough were connected to CF70 ports. All vacuum compo-nents were helium leak tested and the leak rate was found tobe less than 5�10−10 mbar l / s. The objective of the vacuumsystem was to evacuate the plasma chamber, beam line, andits components, to avoid loss of particles due to recombina-tion, and to avoid the electrical breakdown due to high volt-age. The ultimate vacuum requirement is less than 1�10−6 mbar. Turbo molecular pump, with the capacity of400 l / s, was selected by considering gas throughput ��1.5�10−3 mbar l / s� to maintain pressure of the order of10−4–10−3 mbar in plasma chamber, 10−5–10−6 mbar in ex-traction chamber, and to get hydrocarbon free clean vacuum.The whole source was installed on a nonmagnetic supportwith wheel carriages for ease of installation and mainte-a�Electronic mail: skjain@cat.ernet.in

REVIEW OF SCIENTIFIC INSTRUMENTS 78, 053301 �2007�

0034-6748/2007/78�5�/053301/6/$23.00 © 2007 American Institute of Physics78, 053301-1

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:

117.240.185.82 On: Thu, 12 Jun 2014 09:57:43

nance. A gas inlet system consisting of a high purity gascylinder, a pressure regulator, precision standard leak valve,and a flow controller was installed.

The plasma chamber and extraction geometry wereplaced into a special polypropilen tube �dielectric strength�40 kV/mm and operating maximum temperature 100° C�to isolate the solenoid coil from the source body. The extrac-tion geometry was connected with the plasma chamberflange. The ground flange was connected to the vacuumcross-type chamber, having standard rotatable CF203 flange.The alumina flange was used to isolate them.

B. Microwave system

The microwave system has been designed and indig-enously developed, using WR-284 waveguide section. Thedetails of microwave system are presented in Ref. 19. Themicrowave system was sourced by a variable power source�magnetron 2 kW� at a frequency of 2450 MHz. An isolator�isolation 25 dB� was used to protect the magnetron fromload �plasma� reflection. The triple stub tuner was used forwaveguide to plasma impedance matching. The power wasmonitored using a 50 dB directional coupler with detectordiode through calibrated analog meter. The quartz windowprovides vacuum isolation to plasma chamber. The dc breakwas used for the isolation of the microwave system to theplasma chamber as plasma chamber is floated at high poten-tial.

C. Magnetic field

The magnetic field required to satisfy ECR resonancecondition20 is given by f =2.8B �from electron cyclotron fre-quency �ce=eB /m, where, f =microwave frequency, �MHz�,B=critical magnetic field �G�, m=mass of an electron �kg�,and e=electronic charge �C�. The resonant magnetic fieldcorresponding to microwave frequency of 2450 MHz is875 G. The magnetic field in the plasma chamber was ana-lyzed by using POISSON �Ref. 21� software. The electromag-net coils placed around the plasma chamber produces thenecessary magnetic field for ECR resonance condition. Thetwo side coils are identical and center coils are different innumber of turns to produce uniform flat magnetic field con-figuration. The axial magnetic field distribution along theaxis of the source is shown in Fig. 2. The solenoid coil de-sign parameters are presented in Table II.

D. Plasma diagnostics device

The Langmuir probe �tungsten wire of 0.3 mm in diam-eter and 5 mm in length� was used as a plasma diagnosticsdevice for measurement of plasma parameters, plasma den-sity, and electron temperature. During the plasma character-

FIG. 1. The schematic diagram of ECR proton source with 90° mass analyzing magnet.

TABLE I. The design parameters of the ECR source.

Beam energy 50 keVBeam current 30 mABeam emittance �rms normalized� �0.2� mm mradParticles Proton �H+�Frequency 2450 MHzMicrowave power 2 kW �cw�Microwave feed Waveguide �WR-284�Gas pressure base 1�10−6 mbarECR magnetic field 875 GOperating pressure 1�10−4 mbarGas Hydrogen

FIG. 2. The axial magnetic field distribution along the axis of the source.

053301-2 Jain et al. Rev. Sci. Instrum. 78, 053301 �2007�

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:

117.240.185.82 On: Thu, 12 Jun 2014 09:57:43

ization experiments, the probe was inserted axially and thetip of the probe was placed at the center of the last coil justtouching the edge of the plasma surface. The plasma param-eters were determined from an analysis of probe current-voltage characteristics, which was obtained by varying probebias �−150 to +50 V dc� around plasma potential and mea-suring current collected by the probe from the plasma.22–24

The plasma density and electron temperature were computedusing ion saturation and Lam’s method.25–27 Lam’s method isvalid for the unmagnetized plasma, whereas the experimentsare conducted in dc magnetic field. The electrons are as-sumed in thermal equilibrium and the ion Larmor radius ismuch larger than the probe radius; the unmagnetized plasmaprobe theories for ion collection can be used for magnetizedplasma as described in Lam’s theory.

III. EXPERIMENTAL PROCEDURE

A base pressure of 2�10−6 mbar was maintained in theplasma chamber for several hours using turbo molecularpump, prior to the commencement of any experiment. Thepressure was measured using a cold cathode gauge. The sys-tem was then maintained at a pressure �10−4 mbar for about15 min. Gas flow was controlled by precision standard leakvalve. Experimental gas pressure was set to the requiredvalue. The currents in solenoid coils were then adjusted toobtain the desired magnetic field. The microwave generatorwas then switched on for initiating discharge. The micro-wave power �350 W� was coupled to the plasma chamber.Adjusting coil currents a little stabilizes the discharge, fol-lowing which reflected power was minimized using triplestub tuner.

IV. BEAM EXTRACTION ELECTRODE GEOMETRY

The two-electrode extraction geometry for10 mA/15 keV and three-electrode extraction geometry�acceleration-deceleration� for 30 mA/50 keV beam energywas designed to extract the proton ion beam current from thesource. The three-electrode extraction geometry is similar tothat of Taylor and Wills,28 except the diameter of electrodes.This was due to difference in inner diameter of the solenoidcoils. In the geometry of Taylor and Wills electrodes arefixed while in our case the gap in the geometry is adjustable.The overall length and external diameter of the geometry are200 and 250 mm, respectively. Care was taken at the time of

assembly so that the beam axis and extraction geometry axisremain coaxial within precision of 0.1 mm. The gap betweenthe electrodes is very critical, since it governs beam currentand breakdown conditions. The gap between electrodes isadjustable in order to match the plasma meniscus with theextraction geometry at a fixed extraction voltage. This can bedone by breaking the vacuum.

TABLE II. The solenoid coil design parameters �N=total number of turns,N /L=number of turns per layer, and L=number of layer�.

Description Coil 1 Coil 2 Coil 3

Coil type water cooled solenoid coilsCoil size �300�80 �300�70 �300�80

Bore diameter �150 mmNI �ampere turn� 12950 8250 12950Conductor size 5�5��3 copper

N, N /L, L 144,12,12 110,10,11 144,12,12Total conductor length �m� 95 80 95

Power supply 0–32 V, 100 A dc�3 numbersCoil resistance �measured� ��� 0.146 0.127 0.156

FIG. 3. The behavior of rms emittance for three-electrode extractiongeometry.

053301-3 ECR proton source Rev. Sci. Instrum. 78, 053301 �2007�

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:

117.240.185.82 On: Thu, 12 Jun 2014 09:57:43

The primary parameters for the extraction electrode ge-ometry were calculated using standard empiricalrelations.29–31 These parameters were used for input to elec-tromagnetic software IGUN �Ref. 32� to simulate the extrac-tion of positive ions from the plasma. The space charge com-pensation of the extracted beam is taken into account by thesoftware itself. The simulation of two-electrode extractiongeometry for 10 mA/15 keV and three-electrode extractiongeometry �acceleration-deceleration� for 30 mA/50 keVbeam energy was carried out. The simulation results for two-electrode extraction geometry are not presented here due itslimitations towards beam optics. The behavior of rms emit-tance for three-electrode extraction geometry is shown inFig. 3. Fig. 3�a� shows the variation of the rms emittancewith distance from the plasma electrode at different accelera-tion potentials. It is observed from the figure that rms emit-tance decreases with increasing acceleration potential. Figure3�b� shows the variation of rms emittance with ion tempera-ture at fixed acceleration potential and distance from plasmaelectrode. There is a large variation in emittance as distancefrom the plasma electrode is increased; the ion temperature isalso increased due to integration of nonlinear radial electricfield components along the particle trajectories. The gap be-

tween the electrodes is very important which governs thebeam optics. Figure 3�c� clearly shows that at certain gap theemittance is good and it is quadratically increasing. Thus it isvery important to optimize the gap between the electrodes.The trends in the figures are the same as expected with thestandard empirical relations.29–31 The schematic diagram ofthe two and three electrode extraction geometry is shown inFig. 4. The geometrical parameters of the two and three elec-trode extraction geometries are shown in Table III.

V. EXPERIMENTAL RESULTS AND DISCUSSION

A. Plasma characterization

The experiment was aimed at verification of high-density plasma production with the variation of magneticfield, gas pressure, and to match ECR resonance condition at�ce��rf in the plasma chamber. The results that are obtainedwith Langmuir probe at different magnetic fields, gas pres-sures, and transversal variations are discussed below. Theerror associated with the Langmuir probe measurement is±15%.

1. Magnetic field variationsThe magnetic field variation �variation of solenoid cur-

rent� from 45 to 70 A at 2�10−4 mbar pressure, at 350 Wpower, and fixed transversal position �R=0� were studied.The variation of plasma density and hot and cold electrontemperatures with currents, to the solenoid coils are pre-sented in Fig. 5. It is observed from the figure that the sourceis able to sustain high-density plasma over a wide range ofmagnetic field even at low power. The plasma density in-creases with magnetic field due to improved confinement ofthe plasma. Within the error of 15%, the hot electron tem-perature is independent of the solenoid current and it has apeak at 65 A.

2. Pressure variationsThe pressure variations were studied in the range of 2

�10−5–5�10−4 mbar of nitrogen gas at the same micro-wave power of 350 W and same magnetic field at 50 A. The

FIG. 4. The schematic diagram of the two- and three-electrode extractiongeometry.

TABLE III. The geometrical parameters of the two and three electrode extraction geometries.

Parameter

Two-electrodeextractiongeometry Three-electrode extraction geometry

Plasma electrode �PE� �8 mm �5 mm, fixed with respect to PCAcceleration electrode �AE� Absent �10 mmGround electrode �GE� �10 mm �8 mmGap 1 �see Fig. 4� 5 mm,

adjustable5 mm, adjustable up to 20 mm

Gap 2 �see Fig. 4� Absent 3 mm, adjustable up to 20 mmMaterial SS 304 L Molybdenum and copperInsulator Teflon AluminaCooling Air cooled Water cooled, conductivity�1�S/cm, pressure: 6 bars,

flow rate: 3 l /minPower supply for source potential: 50 kV, 100 mA, 3 phase, ripple�0.2%

Power supply for acceleration potential: −3 kV, 20 mA, ripple�0.05%

053301-4 Jain et al. Rev. Sci. Instrum. 78, 053301 �2007�

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:

117.240.185.82 On: Thu, 12 Jun 2014 09:57:43

variation of plasma density and hot and cold electron tem-peratures with gas pressure are presented in Fig. 6. It is ob-served from the figure that plasma density is very sensitiveover a wide range of pressure and there is an enhancement inthe plasma density by a approximately a factor of 2 as com-pared to high pressure range.

3. Transversal variationsThe variation of transverse positions with constant mag-

netic field 50 A at 2�10−4 mbar was also studied. Thevariation of plasma density and hot and cold electron tem-peratures with variation of transversal distance �R=0� arepresented in Fig. 7. It is observed from the figure that plasmais more in the center. The density decreases marginally fromits on-axis value of 3.4�1011–2.6�1011 cm−3 at a radius of±40 mm. The electron temperature shows a broad peakedprofile of 85 eV at a radius of ±40 mm.

It is observed from all the figures that ECR plasma hasmultiple temperature distribution and similar results were ob-served in other ECR plasma source.33–36 The variation in thecold electron temperature is almost negligible. This explains

why the impact ionization cross section for low energy elec-trons ��10 eV� is negligible and is confirmed as standarddata for nitrogen.37

B. Beam characterization

In the initial experiments the total ion beam current of2.5 mA was extracted at 15 keV beam energy, using two-electrode extraction geometry and it is space charge limited.It is four times less than simulation results due to no focusingelements were used here. A cone shaped Faraday cup withcopper target, as a beam measuring device, was used forcollecting the ions. The Faraday cup was placed 300 mmaway from the ground electrode. The measured total ionbeam current as a function of extraction voltage is shown inFig. 8. The aerial view in the figure shows the impact ofbeam on Faraday cup. The testing with three-electrode ex-traction geometry �30 mA/50 keV� is in progress. The de-sign and fabrication of 90° mass analyzing magnet to studythe proton fraction in the beam is complete. The specifica-tions of the analyzing magnet are shown in Table IV. Thesupervisory control system for monitoring the various pa-

FIG. 5. The variation of plasma density and hot and cold electron tempera-tures with current to the solenoid coils.

FIG. 6. The variation of plasma density and hot and cold electron tempera-tures with gas pressure.

FIG. 7. The variation of plasma density and hot and cold electron tempera-tures with variation of transversal distance �R=0�.

FIG. 8. The measured total ion beam current as a function of extractionvoltage.

053301-5 ECR proton source Rev. Sci. Instrum. 78, 053301 �2007�

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:

117.240.185.82 On: Thu, 12 Jun 2014 09:57:43

rameters is under development similar to Ref. 38. The sole-noid based low energy transport line can be used to transportproton beam directly to radio frequency quadrupole formatching.

ACKNOWLEDGMENTS

The authors wish to thank Mrs. V. Sekar and Mr. DheerajVerma for their help and kind support for assembly and test-ing of electronic circuits, data acquisition system, and testingof magnetron power supply and Mr. R.J. Parmar for systemsdrawings. The authors also like to thank Mr. R. K. Mishraand Mr. Ritesh Malik for the design and development of thesector magnet. The staff members of accelerator workshopand chemical section are acknowledged for their work offabrication and chemical cleaning and electropolishing ofcomponents in time.

1 R. Geller, B. Jacquot, and M. Pontonnier, Rev. Sci. Instrum. 56, 1505�1985�.

2 C. Lyneis, Nucl. Instrum. Methods Phys. Res. B 10/11, 775 �1985�.3 B. Chapman, Glow Discharge Process �Wiley, New York, 1980�.4 T. Ono, C. Takahashi, and Seitaro Matsuo, Jpn. J. Appl. Phys., Part 2 23,L534 �1984�.

5 Z. Sitar, M. J. Paisley, D. K. Smith, and R. F. Dawis, Rev. Sci. Instrum.61, 2407 �1990�.

6 K. Kretshmer, K. Matl, G. Lorenz, and I. Kessler, Solid State Technol. 33,53 �1980�.

7 N. Sakudo, Nucl. Instrum. Methods Phys. Res. B 21, 168 �1987�.8 T. Taylor and J. S. C. Wills, Nucl. Instrum. Methods Phys. Res. B 30, 37

�1991�.9 Z. Song, S. Peng, J. Yu, J. Ming, Z. Yuan, F. Qian, and Z. Guo, Rev. Sci.Instrum. 77, 03A305 �2006�.

10 T. Taylor and J. Mouis, Nucl. Instrum. Methods Phys. Res. A 336, 1�1993�.

11 R. R. Stevens Jr., Sherman, and J. D. Schneider, Proceedings of the 1993Particle Accelerator Conference, Washington �unpublished�.

12 A. Farchi, M. Delaunay, G. Mellin, R. Gobin, P. A. Leroy, J. Faure, O.Delferriere, and J. M. Lagniel, Rev. Sci. Instrum. 67, 1240 �1996�.

13 J. S. C. Wills, R. A. Lewis, J. Diserens, H. Schmeing, and T. Taylor, Rev.Sci. Instrum. 69, 65 �1998�.

14 J. Sherman et al., Rev. Sci. Instrum. 73, 917 �2002�.15 J. Sherman et al., Rev. Sci. Instrum. 69, 1003 �1998�.16 R. Gobin et al., Proceedings of EPAC 2002, Paris, France �unpublished�.17 S. Gammino, G. Ciavola, L. Celona, L. Torrisi, D. Mascali, S. Passarello,

and A. Galata, Rev. Sci. Instrum. 77, 03B511 �2006�.18 D. D. Bhawalkar et al., Proceedings of the Indian Particle Accelerator

Conference, Indore, India, 2003 �unpublished�.19 S. K. Jain, A. Jain, and P. R. Hannurkar, Indian J. Pure Appl. Phys. 42,

896 �2004�.20 I. G. Brown, The Physics and Technology of Ion Sources �Wiley, New

York, 1989�.21

POISSON, LANL code, Reference Manual LA-UR-87-126, 1987.22 I. M. Hutchinson, Principles of Plasma Diagnostics �Cambridge Univer-

sity Press, Cambridge, 1987�.23 N. Hershkowitz, Plasma Diagnostics �Academic, New York, 1989�.24 K. E. Andersson, M. K. Wahlstrom, and A. Roos, Thin Solid Films 214,

213 �1992�.25 S. H. Lam, Phys. Fluids 8, 73 �1963�.26 F. F. Chen, R. H. Huddlestone, and S. L. Leonard, Plasma Diagnostics

Techniques �Academic, New York, 1965�.27 S. K. Jain, A. Jain, D. Sharma, and P. R. Hannurkar, Indian J. Phys., A 80,

1011 �2006�.28 T. Taylor and J. S. C. Wills, Nucl. Instrum. Methods Phys. Res. A 309, 37

�1991�.29 R. Keller, Nucl. Instrum. Methods Phys. Res. A 298, 247 �1990�.30 T. Taylor, Rev. Sci. Instrum. 63, 2507 �1992�.31 H. Zhang, Ion Sources �Science, New York, 1999�.32 R. Becker, IGUN, Germany.33 Y. Kawai and K. Sakamoto, Rev. Sci. Instrum. 53, 606 �1982�.34 A. Gangui, R. Baskaran, and H. D. Pandey, IEEE Trans. Plasma Sci. 18,

134 �1990�.35 R. Baskaran, Ph.D. Thesis, IIT Delhi, 1990.36 R. Baskaran, S. K. Jain, and S. S. Ramamurthi, Rev. Sci. Instrum. 63,

2525 �1992�.37 A. Von Engel, Ionized Gases �Clarendon, Oxford, 1955�.38 P. Kumar, G. Rodrigues, U. K. Rao, C. P. Safvan, D. Kanjilal, and A. Roy,

Pramana, J. Phys. 59, 805 �2002�.

TABLE IV. The specifications of the analyzing magnet.

Magnet-C type Parameters

Bending angle 90°Good field region ±30 mm

Bending radius 120 mmDesign field 3000 G

Homogeneity 1�10−3

Pole gap height/width 26 mm/112 mmPower supply 25 V, 15 A dc

053301-6 Jain et al. Rev. Sci. Instrum. 78, 053301 �2007�

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:

117.240.185.82 On: Thu, 12 Jun 2014 09:57:43