calibration experiment of all-sky electrostatic analyzer...178 giott [9], ulysess [10], and galileo...

1 Introduction

Electrostatic analyzers have commonlybeen used as plasma measuring instrumentsaboard satellites to measure the energy distri-bution and three-dimensional velocity distri-bution of space plasma. The electrostatic ana-lyzer measures the E/q (energy per charge) ofparticles using the unique kinetic characteris-tic of charged particles in an electrostatic field.The measured energy of space plasma (ionsand electrons) around a satellite and the inci-dent angle distribution obtained by the analyz-er are used to calculate the plasma flow veloc-ity, density, and temperature and to detectlarge-scale space disturbances. The informa-tion obtained on the plasma structure andpropagation is used to forecast space weather.Further, the micro-scale velocity distributionfunctions of ions and electrons provide us witha better understanding of particle acceleration,wave-particle interactions, and the propaga-tion of accelerated particles in space plasma.Moreover, by adding [1] a TOF (Time-Of-Flight) velocity analysis function to the final

stage of the electrostatic analyzer, it will bepossible to identify individual species of ions(such as H+, He++, and O+).

In an electrostatic analyzer, two opposingelectrodes generate a potential differencethrough which charged particles are passed.These analyzers are categorized in accordancewith the structure of the electrodes, andinclude the parallel plate type, the concentriccylinder type, and the spherical type. Thespherical electrostatic analyzer has only oneinlet, which particles enter from variousangles. Since the exit position of a particlevaries according to its incident angle, detect-ing the exit locations enables the simultaneousmeasurement of the incident angle distribution(one-dimensional) and the energy distribution.Therefore, when a satellite mounted with aspherical electrostatic analyzer rotates aboutits spin axis, the analyzer is able to performthree-dimensional velocity distribution meas-urement. Because of this advantage, sphericalelectrostatic analyzers are often installed onsatellites for space plasma measurement. Todate, the ISEE [2]-[5], GEOS [6], AMPTE [7][8],

MIYAKE Wataru and YAMAZAKI Atsushi 177

Calibration Experiment of All-sky Electrostatic AnalyzerMIYAKE Wataru and YAMAZAKI Atsushi

Electrostatic analyzers have been widely used for measuring the energy and incidentangle distribution of space plasma on board the spacecraft. The measurement enables us toderive the flow velocity, density, and temperature of plasma, to detect the large-scale distur-bances in space, and to forecast the space weather. An electrostatic analyzer with all-skyfield of view has been developed for the L5mission and the proto-type model is calibratedby a ground facility. The results demonstrate a fair agreement with the numerical model cal-culations and its capability of the space plasma measurement on board 3-axis stabilizedspacecraft such as the L5 mission.

Keywords L5 mission, Solar wind plasma, Electrostatic analyzer

178

Giott [9], Ulysess [10], and Galileo [11] havecarried onboard spherical electrostatic analyz-ers. However, even a spherical analyzer with awide angle of visibility still has a range ofdead angles, and its analytical characteristicwill therefore vary depending on the line ofsight. This drawback was solved in the so-called “top-hat” electrostatic analyzer [12][13].Its axially perfect symmetrical shape ensuresuniform analytical characteristic over a fieldof view of 2πrad.

In Japan, three analyzers were designedfor the observation of solar wind plasmas andmounted on satellites. Two were 270°spheri-cal electrostatic analyzers [14][15], and thethird was a top-hat electrostatic analyzer [16].However, these analyzers were all mounted onsatellites with spin-stabilized attitude controland thus featured only one-dimensional angu-lar resolution. By positioning the one-dimen-sional field of view on a plane that partiallyincludes the satellite’s rotation axis, the satel-lite’s rotation and the one-dimensional angularresolution are combined to enable the meas-urement of angle distributions of two orthogo-nal components, and also to provide a three-dimensional velocity distribution functionwhen coupled with measurement of energydistribution.

With a satellite using three-axis attitudecontrol, in theory it is not possible to performthree-dimensional velocity distribution meas-urement using the above-mentioned three ana-lyzers. Three-axis control is expected to beadopted for the L5 mission [17] aimed at imag-ing CME in interplanetary space and activityon the sun’s surface. In such a satellite, inno-vative measures are required to widen the fieldof view of the measuring instrument. Onemethod of establishing such an electrostaticsweep is to install electrostatic deflecting elec-trodes at the entrance section of the analyzer.In the WIND satellite, the collimator sectionof the onboard top-hat electrostatic analyzerwas provided with deflecting electrodes toachieve a field of view greater than ±45degrees [18]. Figure 1 (a) is a schematic dia-gram of this type of analyzer. The arrows in

the diagram indicate the traveling directions ofparticles that enter and exit the instrument.The field of view is the circular area ofapproximately ±45 degrees from both sides ofthe plane (indicated by a dotted line in the dia-gram) perpendicular to the analyzer’s symme-try axis. As a result, the satellite body can eas-ily interfere with the analyzer’s field of view.Thus for maximum performance, this type ofanalyzer must be installed at the tip of anextended element, such as a boom.

On the other hand, a bottle-type analyzeroffers a full-sky field of view (2π str), con-ducting its sweep by opening and closing aconical surface with a field of view formed atan angle of 45 degrees from the analyzer axis.Fig.1 (b) is a schematic diagram of this type ofanalyzer. Unlike the top-hat analyzer, this ana-

Journal of the National Institute of Information and Communications Technology Vol.51 Nos.1/2 2004

Cross-sectional diagram andschematic of incident and exit parti-cles of top-hat electrostatic analyzer(a) and newly developed electrostat-ic analyzer prototype (b)

Fig.1

lyzer can be mounted directly on the outer sur-face of a satellite, a major advantage.

To examine this analyzer in detail, weobtained a potential distribution based on spe-cific shape and electrode arrangement, andused a numerical model to calculate the travel-ing paths of a large number of particles withinthe analyzer [19]. This investigation revealedfavorable analyzer performance. Accordingly,this paper introduces a prototype electrostaticanalyzer with a full-sky field-of-view sweepfunction, designed for installation on a satel-lite such as the L5 mission featuring three-axiscontrol, and also discusses the results of anexperiment we conducted to evaluate basic

analyzer performance, using a specific ion-beam charged-particle calibration device [20].

2 Outline of the prototype analyz-er

A cross section of the prototype analyzer isshown in Fig.2. In this figure, the prototypeanalyzer has been rotated 90 degrees from thatshown in Fig.1 (b). The analyzer is mademainly of aluminum and constitutes the con-ductor section (blue shaded area in the dia-gram), and the deflection plates and energyanalyzer electrodes are electrically insulatedby UPIMOL (red shaded area in the diagram).


Cross-sectional view of prototype analyzerFig.2

180

Since the prototype was produced to testthe analyzer’s basic functions in a groundfacility, we designed it without taking weight,impact resistance, other mechanical character-istic, or thermal characteristic into considera-tion. We also determined the shape with amain focus on achieving dimensional preci-sion without the use of special assembly tech-niques. We ensured an accuracy of 5 ±0.05mm for the passage width along the chargedparticle channels. The shape of this prototypeanalyzer was determined based on the resultsof numerical calculation, and the size [19] wasoptimized for the measurement of solar windelectrons aboard the L5 mission. Althoughsome have expressed the opinion that a largerdevice would be more suitable for measure-ment of solar wind ions (this will be discussedlater in detail), we determined the size of theprototype component in view of the uniformdiameter of the ion beams generated by thecharged-particle calibration device [20] used inour experiment. A larger instrument wouldrender quantitative evaluation difficult. Inother words, the prototype was produced toassess basic performance with respect to theanalyzer shape; the optimal size will be inves-tigated based on these results and in accor-dance with the scaling law of charged particlepaths in an electrostatic potential field.

The prototype analyzer consists primarilyof three types of electrodes: electrostaticdeflecting electrodes for line-of-sight scan-ning, limiter electrodes to adjust sensitivity,and spherical electrodes to analyze particleenergy. Electrostatic potentials created by anexternal power supply are applied to the elec-trodes. The analyzer is in cylindrical symme-try. Particles enter the analyzer from the upperleft and lower left and move toward the centerof the X shape in the diagram. Particles passthrough a mesh section that prevents the ana-lyzer’s internal potential from leaking to theoutside. Particles then travel between the twofacing deflecting electrodes each featuring acuneiform cross section. By applying a poten-tial to these facing deflection electrodes, theanalyzer generates an electrical field between

the electrodes and bends the paths of incidentparticles to freely deflect the line of sight.

After particles pass the center of the X,they travel past the thin annular limiter elec-trodes, which serve as the “diaphragm” forsensitivity adjustment. Application of a volt-age to the limiter electrodes reduces the num-ber of passing particles, enabling the measure-ment of a high particle flux. Since the dynam-ic range (number of particles measured pergiven period) of the particle detector in thefinal stage is limited, this function was adopt-ed to expand the dynamic range of the meas-uring instrument and to enable measurementof both extremely small and very large particleflux.

Particles are then led to the spherical elec-trostatic analyzer section. The radii of theouter and inner spheres are 52.5 mm and 47.5mm, respectively. A potential difference isapplied to the space between the two spheres,so particles travel in a near-circular orbit. Par-ticles with energies that are too great relativeto the applied potential difference travelalmost linearly, strike the outer sphere, and areunable to exit. Orbits of particles with verylow energies are bent substantially; these par-ticles collide with the inner sphere, and aresimilarly blocked. As a result, only particles atthe appropriate energy level exit the sphericalanalyzer.

Particles passing through the sphericalanalyzer reach the exit located at the right end,where the detector converts them to electricsignals, which are detected as a pulse count. Inour experiment we used a microchannel plate(MCP) [21], a type of secondary electronamplifying tube, as a detector. Particles flyaround the analyzer in all directions and at allenergy levels, but only particles with incidentangles and energy levels satisfying certainconditions relative to the potential applied tothe electrodes reach the final-stage detector.All other charged particles collide against thepassage walls before reaching the final stage.Changing the combination of voltages appliedto the individual electrodes and examining thenumber of passing particles provides informa-


tion on the angle-energy distribution of thecharged particles within the analyzer. For fur-ther details, refer to the investigation based onthe numerical model calculations [19].

3 Outline of test equipment andtest procedures

Prototype performance was investigatedwith a charged-particle calibration device [20].This device produced parallel N2

+ ion beams ata uniform energy level within a vacuum cham-ber, albeit in conditions quite different fromthose of actual plasma in space. Since the ionbeam generator was a particularly elaboratedevice, we did not vary its energy level ordirection of radiation, but instead maintainedthese at constant values. Instead, to studyangular characteristic, the analyzer was placedon a vacuum gimbals (rotating platform) withtwo orthogonal axes, and was rotated to varythe direction at which the beam entered theanalyzer. To investigate energy characteristic,on the other hand, the voltages applied to theelectrodes inside the analyzer were varied infine increments. Figure 3 shows the analyzersetup on the dual-axis vacuum gimbals insidethe vacuum chamber. Ions were directed fromleft to right in the photo.

All devices used to measure analyzer char-acteristic were controlled from a single per-

sonal computer via GPIB communication.These devices included the following:* Analog voltage power supply unit, providingcontrol voltage to the ion-beam control equip-ment* Multiplexer and multimeter to monitor ionbeam current and voltage* Controller for gimbals control* Multiplexer and multimeter to monitorpower applied to the analyzer and output volt-age* Detector output pulse counter* Ionization vacuum gauge (for monitoring ofdegree of vacuum)

Figure 4 shows the control screen dis-played on the PC. To enable independent con-trol, we included separate control panels (onthe right side of the screen) within the applica-tion. The sample screen shows the main panelfor automatic control sweep measurement andthe sweep control panel. To check the analyzercharacteristic for a selected parameter, allother parameters were fixed, and an automaticsweep was conducted for the selected parame-ter while output pulse number was tallied,enabling the smooth acquisition of data. Thesweep measurement items can be dividedroughly into the following three groups:(1) Measurements related to output ion beams(voltage/current sweep)(2) Measurements related to analyzer direction(incident angle sweep)(3) Measurements of voltage applied to ana-lyzer (energy, angle of visibility, sensitivitysweep)

Measurements of the first type were con-ducted by applying a control voltage to the ionbeam control device from the analog voltageoutput device controlled by the GPIB commu-nication. For type 2 measurement, a sweepwas conducted by rotating the gimbals insidethe chamber. For type 3 measurement, thevoltage applied to the observation equipmentwas swept while the power supply voltage wascontrolled directly by the GPIB.

In the experiment, the control panels (seeFig.4) on the PC display were operatedaccording to the flow chart shown in Fig.5.


Prototype analyzer on dual-axis vacu-um gimbals (rotating platform) insidevacuum chamber

Fig.3

182

First, a sweep item was selected, and all otherparameters were fixed using the main panel.

Next, using the sweep control panel, theinitial value, final value, and step value, aswell as the observation time at each measuringpoint, were set. The Scan Start button wasclicked to start sweep measurement. At eachpoint, measurement began when the devicestabilized, one second after termination of theparameter sweep. After the measurement wascompleted at each point, the result was outputand displayed in the space below the sweepcontrol panel. After the measurement wascompleted at the final sweep point, the datawas output as a file for analysis.

4 Analyzer characteristic

4.1 Energy-angle characteristicSince the analyzer was axially symmetri-

cal, dependency on elongation of the axisshould be investigated. Here, the in-planeorthogonal to this axis was defined as havingan elevation angle of 0 degrees, and the left-ward plane parallel to the axis in Fig.2 was setas 90 degrees. When the deflecting electrodesfor the field-of view sweep were grounded(i.e., when the deflection was 0), incident par-ticles would reach the entrance of the spheri-cal energy analyzer centering around the ele-vation angle of 45 degrees. By rotating andfixing the gimbals at a certain elevation angle

and incrementally varying the potential(sweep voltage, or SV) applied to the innerand outer spherical electrodes of the sphericalanalyzer, we were able to count the number oftransmitted particles per unit time in each step.In this experiment, a positive potential wasapplied to the outer sphere, and a negativepotential of the same degree was applied tothe inner sphere. Next, we moved the gimbalsslightly to change its fixed elevation angle andconducted the SV sweep again over the samerange. Repeating this process yielded the ener-gy-angle characteristic near 45 degrees shownin Fig.6. In the graph, the maximum transmis-sion rate percentage (particle count) wasassigned the value of 100%, contour lineswere also plotted at 80%, 60%, 40%, and20%.

Charged particles featuring the same ratioof kinetic to potential energy (i.e., the samepotential distribution) followed the same pathsin the electrostatic field, in accordance withthe scaling law. Therefore, the degree of parti-cle energy relative to SV was indicated on the


Control panels on PC screenFig.4

Measurement flow chartFig.5

horizontal axis. When ±100 V was applied toboth the outer and inner spheres, particles withan energy level of 850 eV to 1,000 eV passedthrough. This SV value is proportionally relat-ed to the energy value of transmitted particles.In other words, particles of approximately 9 *SV [eV] relative to the applied ±SV will passthrough the analyzer and be detected by thedetector.

The energy resolution and angular resolu-tion are defined as half-value widths of theintegral distribution in the vertical and hori-zontal directions, as seen in Fig.6. Since theabove scaling law applies to the horizontalaxis, the resolutions can be defined as the ratioto the median of the half-value widths. In thiscase, the energy resolution is approximately±7%, and the angle resolution is approximate-ly ±1.8 degrees.

Figure 7 shows the energy-angle charac-teristic of the analyzer obtained based onnumerical model calculations [19]. This char-acteristic is expressed in the same format asthat used in Fig.6, and agrees well to the graphin Fig.6, indicating that this characteristic wasas estimated using the numerical calculations.Detailed comparison, however, revealed aslightly wider spread in both energy and angle

distributions. We suspect that this was attribut-able to limitations posed by the singularity ofenergy and the parallelism of the ion beamsused in the experiment, in addition to manu-facturing and assembly deviations, but wecould not determine the exact cause. Never-theless, we believe that the close correspon-dence between predicted and actual resultssupports the practical application of the ana-lyzer under study.

The absolute sensitivity of an analyzer isreferred to as the G factor, and it indicates theratio of transmission energy width multipliedby the transmission solid angle to the productof opening area and center transmission ener-gy. In other words, integration of the transmis-sion counts indicated by the contour lines inFigs. 6 and 7 by energy and angle, in additionto integration of the directions orthogonal tothe elevation angles, will yield the G factor,provided that the incident particle flux isknown. The number of particles C (/sec) perunit time measured at the center transmissionenergy E and the differential flux F (/cm2 strkeV sec) displays the following relation: C =F(E)●G●E●ε. Here,εis the detection efficien-cy of the detector and G is the G factor. In ourexperiment, we could not measure the


Energy-angle characteristic (meas-ured value) at an elevation angleapproaching 45 degrees

Fig.6Energy-angle characteristic (modelcalculation) at an elevation angleapproaching 45 degrees

Fig.7

184

absolute ion beam flux with sufficiently highprecision; therefore, we were unable to obtainthe G factor. However, since the energy-anglecharacteristic matches that of the numericalmodel calculations, it is safe to assume theapplicability of the G factor value obtained inthese calculations.

In both Figs. 6 and 7, coupling is seenbetween particle energy/SV and elevationangle (EL), and skewing [22] (distortion) isgenerated. This skewing occurs when thebending angle (angle of deflection of particlepaths inside the spherical analyzer) is smallerthan 180 degrees [23]. Particles with highenergies are not easily deflected and thus theyare able to reach the exits near the outersphere. Therefore, when the incident angle isdirected outward (from the inner spheretoward the outer sphere), particles move evenfarther away from the centers of the concentricspheres at the exits, collide against the outersphere, and are blocked from passing throughthe analyzer. If the incident angle is directedinward, on the other hand, particles tend todeflect toward the center (toward the innersphere), thus mitigating the effect of theirhigher energies and allowing them to passthrough. With particles at lower energies, theopposite phenomena are seen. As a result,skewing is generated between energy and inci-dent angle, as shown in Figs. 6 and 7.

When processing measurement data foractual space plasma, this skewing and slightvariation in transmission rate can be ignoredin most cases. In other words, in the case of ananalyzer with an energy resolution ofδ% andangular resolution ofαdegrees, uniform trans-mission is assumed in the area within ±δ%relative to the center energy corresponding tothe applied SV and within ±α degrees fromthe direction in which the analyzer faces. Thisarea corresponds to that of the rectangle inFigs. 6 and 7. Outside this area, transmissionis considered to be zero. This approximation isvalid when the energy-angle distribution ofmeasured plasma is sufficiently large in rela-tion toδ andα. For example, when the tem-perature of plasma is extremely low, the ana-

lyzer must have low values forδandα in orderto obtain accurate measurements.

4.2 Field-of-view deflectionWe examined the relationship between the

deflection voltage (Vdef) and elevation angle(El. angle) observed when an electric potentialis applied to the deflecting electrodes for afield-of-view sweep. Figure 8 shows the resultsof this experiment, and Fig.9 indicates theresults of the calculations using the numericalmodel. As described in (4.1), the particle ener-gy and the voltage applied to the electrodesare governed by the scaling law; the ratio ofdeflection voltage to particle energy is shownon the vertical axis. In the experiment, a posi-tive potential was applied to one of the twodeflection electrodes and the other electrodewas grounded (zero potential). In the diagram,those ratios of the positive side of the verticalaxis indicate the application of a positive volt-age to the electrode, which was near the sym-metry axis and cylindrical in shape (as seen inFig.2), while those ratios of the negative sideof the vertical axis indicate the application ofpositive voltage to the disc-shaped electrode,for easier understanding. This figure shows anoverall view of the field-of-view sweep char-acteristic. At the zero position on the verticalaxis, both electrodes were grounded.


Field-of-view deflection (measuredvalue)

Fig.8

In Fig.8, the contour lines are discontinu-ous, but this was attributable to the large scat-tering of measurement points. In reality, thecontour lines would be smooth. In Fig.9,stepped changes can be seen, but these weredue to the finite number of numerical calcula-tion points. In both graphs, a near-linear rela-tion can be observed between the appliedvalue and field-of-view deflection angle. Fur-thermore, the entire field of view is coveredwith a maximum voltage of 70 % to 80 % ofthe particle energy.

Although the electrostatic deflection elec-trodes appear to be symmetrical with respectto the conical 45-degree surface, a three-dimensional view indicates that the deflectingelectrode near the axis is almost cylindrical inshape, and that the other deflection electrodeis nearly flat. Therefore, the formed potentialstructure acts as a convex mirror when a volt-age is applied to the cylindrical electrode nearthe symmetry axis, and as a concave mirrorwhen a voltage is applied to the disc-shapeddeflection electrode. Therefore, parallel inci-dent particles diffuse in the former case (< 45degrees), and they converge in the latter case(> 45 degrees).

Due to these effects of diffusion and con-vergence, the percentage of transmission ishigher for incident particles with a large eleva-

tion angle than for incident particles with asmall elevation angle. When a voltage wasapplied to the flat electrode (> 45 degrees), thedeflection was slightly higher (same degree ofdeflection observed at a low voltage) thanwhen a voltage was applied to the cylindricalelectrode near the symmetry axis (< 45degrees). We suspect that the slight variationin deflection may be linked to the potentialstructure; i.e., whether it promoted diffusionor convergence.

4.3 Limiter characteristicLast, we investigated the sensitivity

adjustment function by applying voltage to theannular limiter electrodes installed betweenthe field-of-view sweep electrostatic deflect-ing electrodes and the spherical electrodes forenergy analysis. The results are shown inFig.10. The graph shows the integration (sum)over angles in the data obtained in SV sweep-ing at varied incident angles. The solid linecorresponds to the data in Fig.6, and the volt-age applied to the limiter was zero. Thedashed line indicates the data obtained byapplying a voltage equivalent to 80 % of parti-cle energy, while the dotted line shows thedata acquired by applying a voltage equivalentto 90 % of particle energy. Figure 11 showsthe results of the corresponding numericalmodel calculations.

The measured values and the model calcu-lation results matched closely, thus indicatingthe feasibility of controlling an approximatelytenfold variation in sensitivity, as we expect-ed. Application of voltage to the limiter elec-trodes not only reduced the number of trans-mitted particles but also increased the trans-mitted particle energy/SV value slightly, aspredicted under the numerical model. Thisdemonstrated that this type of limiter not onlyhinders the transmission of particles withenergies lower than the applied voltage, butthat it also modifies the paths of the transmit-ted particles, performing its limiting functionin combination with the energy-incident anglecharacteristic of the spherical analyzer in thelater stage. Therefore, even an applied voltage


Field-of-view deflection (model calcu-lation)

Fig.9

186

of approximately 80% of particle energy/qwill result in a noticeable decrease in sensitivi-ty.

Although the limiter electrodes werearranged as if the inside and outside of theelectrodes would provide about the same func-tion and form an equipotential surface nearly

perpendicular to the direction of particlemotion shown in Fig.2, the convergence/diffu-sion lens effects were the same as those seenwith the field-of-view sweep electrostaticdeflecting electrodes, converging particlebeams that were parallel before entering thespherical analyzer to a focal point, as predict-ed particle paths under the numerical model.Due to the three-dimensional effect, the focalpoint is slightly displaced from the center,toward the exterior. Therefore, particles enter-ing the spherical analyzer became diffusedparticle beams emitted from a focal point clos-er to the outer sphere. This means that theincident particles followed a path from theoutside toward the inside of the outer sphere.Therefore, similar to the skewing of the ener-gy-incident angle of the spherical analyzerexplained in (4.1), the transmission of parti-cles directed inward features a lower SV value(that is, a higher particle energy/SV) and mini-mal bending. We believe that this caused theparticle energy/SV to increase slightly, as indi-cated by the dashed and dotted lines in Figs.10 and 11.

5 Summary and discussion

The following summarizes the main char-acteristic of the prototype analyzer.1) The energy resolution and incident angularresolution were ±7 % and ±1.8 degrees,respectively, and the center transmission ener-gy/analyzer applied voltage was 9 (eV/V).2) By applying a maximum voltage of 70 % to80 of the particle energy/q to the electrostaticdeflecting electrodes, the analyzer achieved afield-of-view sweep of ±45 degrees for a full-sky field of view (2π str).3) By applying a voltage of 80 % to 90 of theparticle energy/q to the limiter, it is possible toreduce sensitivity to approximately 10 % ofthe original value.

These values were very close to those esti-mated based on the calculation results of thenumerical model, although slight deviationswere noted. It was thus confirmed that the pro-totype analyzer fulfilled its designed func-


Limiter characteristic (measured val-ues).

Fig.10

Solid, dashed, and dotted lines representdata obtained with the limiter voltage/par-ticle energy values of 0.0, 0.8, and 0.9(V/eV), respectively.

Limiter characteristic (model calcula-tions).

Fig.11

Solid, dashed and dotted lines representdata obtained under the same conditions asthose in Fig.10.

tions. The validity of the numerical modelproduced prior to design was also verified,indicating that the model may be used fordetailed investigation and design analysis inthe future. Although the instrument’s sensitivi-ty (G factor) was not measured directly in ourexperiment, given the fact that the above ener-gy-incident angle characteristic correspondedto the model calculation results, we concludedthat the actual G factor would also correspondto the predicted value under the numericalmodel. Although a number of aspects remainto be examined prior to installation on a satel-lite—such as vibration and impact resistancemechanical characteristic, and thermal charac-teristic in a vacuum—development of our pro-totype represents a major step forward in thedevelopment of a plasma measurement instru-ment with a full-sky field of view to bemounted on a satellite with three-axis control.

Next we will discuss solar wind plasmameasurement in the context of the L5 mission.Solar wind electrons consist of isotropic ther-mal cores of up to approximately 30 eV andhigh-energy tail halos elongated along themagnetic field lines to several hundred eV [24].Further, electrons accelerated during the gen-eration of solar flares propagate at energies ofseveral hundred eV or more, mainly in thedirection of the magnetic field lines [25].

For measurement of thermal cores andhalos of solar wind electrons requiring a fieldof view in nearly all directions (4π str), instal-lation of two measuring instruments on oppo-site sides of a satellite may prove effective.The full-sky field of view is limited from 70 %to 80 % of the maximum output (several keV)of the high-voltage sweep power supply, butmaximum measurement energy can beobtained in the distribution along the 45-degree cone. In the case of electrons, thevelocity distribution is expected to be depend-ent on the pitch angle relative to the magneticfield lines; therefore, when the conical surfaceand the magnetic field lines are parallel, allpitch angles are covered. For installation on asatellite, it is best to position the analyzer sothat the direction of the 45-degree field-of-

view center conical surface is aligned with thedirection of the Parker spiral (45 degrees eastof the sun) toward which interplanetary mag-netic fields are, statistically speaking, mostfrequently pointed. This is expected to coverthe angular distribution of electrons accelerat-ed by high-energy flares.

In terms of angular and energy resolution,electron measurement poses no problem, andcan be used in the measurement of tempera-ture and unisotropy of thermal electrons.Moreover, the numerical flux of the electronsof the solar wind is a maximum of approxi-mately 107/cm2 sec str eV. Assuming that thewidth of the detector anode for azimuth distri-bution measurement is roughly 30 degrees, thesensitivity (G factor) near the incident angle of45 degrees is 7.2×10-4 cm2 str eV. From this,the maximum count is estimated to be 104 persec. The maximum total count over the entiredetector surface is estimated to be 105 per sec.

Solar wind ions consist mainly of protonsin supersonic flow. The particle count is high,and these particles are easy to measure withavailable instrument sensitivities. On the otherhand, thermal velocity (several tens of km/sec)is low in relation to bulk velocity (severalhundred km/sec). This necessitates higherenergy resolution and angular resolution thanrequired to measure ions in the magnetos-phere, if we are to secure sufficient energy andangle distributions to perform the velocity dis-tribution function. Thermal proton measure-ment enables the calculation of basic solarwind macro parameters such as bulk velocity,density, and temperature. These basic parame-ters are used to detect and identify shockwaves, disturbances, CIR, and high-speedstreams resulting from CME in solar winds.The variations of these parameters, combinedwith magnetic field data, are used to detectand identify MHD waves in solar winds, andthey provide the basic information concerningpitch angle scattering, heating, and accelera-tion of ions.

The current energy resolution and angularresolution near an incident angle of 45 degreesare barely sufficient to allow for the proton


188

velocity distribution function. On the otherhand, the proton energy flux reaches an orderof nearly 108/cm2 sec str eV. Therefore, basedon a G factor with an assumed azimuth widthof 1 degree, a count on the order of 104/cm2

sec str eV can be estimated. It is unrealistic toarrange anodes of 1-degree width; widths of 3to 5 degrees are suspected to be the limit. Inaddition, due to the widening angle of protonsin the solar wind, the total count over theentire detector surface is on the order of105/sec.

Thermal protons have an energy level ofapproximately 1 keV. Since the velocity of theproton flow is about the same as that of heavyions in solar winds, observation of this E/qwith an analyzer would result in a distributioncorresponding to the ionic M/q. In otherwords, the peak of He++ appears at the areawhere the E/q is twice as high as that of pro-tons (H+).

Various ions in solar winds not only pro-vide data on the above-mentioned basic solarwind parameters, but also serve two importantpurposes in observation. As these ions consistof high-energy particles generated and accel-erated by shock waves in interplanetary space,their compositions and ionization states offervaluable information on CMEs and other phe-nomena, enabling in-depth investigation ofCMEs and the history of the sun. The highestnumber of heavy ions in the solar winds areHe++ ions (αparticles), but their density is lessthan 20 % of that of solar wind protons [26].The figures for other heavy ions are evenlower. Therefore, the three-dimensional veloc-ity distribution function is required only forprotons and He++. For other heavy ions, all thatis required is an energy distribution based onthe integration of angular arrival directions[27].

Ions in a solar wind flow at supersonicspeed away from the sun. Therefore, they aremore concentrated closer to the sun’s directionand as a result there is no need for a full-skysweep of the electrostatic deflection electrodesfrom 0 degree to 90 degrees; only the areanear the sun needs to be scanned. Therefore,

the analyzer must be positioned in such a waythat the line of sight of the 45-degree conicalsurface is aligned toward the sun. To detectparticles reflected or accelerated by impulsewaves and pick-up ions, the analyzer shouldbe capable of sweeping over certain angles ofvisibility. In our analyzer, when the high-volt-age power supply (maximum output power of3 kV) was used, the angle of visibility for par-ticles of more than 10 keV was limited toapproximately 45 ±15 degrees or less. Sincedata is obtained in the azimuth direction alongthe 45-degree conical surface, a certain rangecan be covered for the measurement of high-energy ions by combining the two of dataobtained.

The gain of the MCP used in a detectordrops when the output current (count) increas-es excessively [21]. Although this saturationcharacteristic depends on the resistance of theMCP, 106/sec is the guideline value. Theexpected lifespan of the MCP is said todepend on the total electric charge dischargedfrom the MCP [21]. Therefore, it is importantto take measures to ensure that the applicationof voltage to the sensitivity adjustment elec-trodes will not result in an input count exceed-ing 105/sec. Currently this requirement isexpected to result in the area near the peak ofprotons. Therefore, a limiter must be used formeasurement in the low energy range, and, inthe high-energy range, the maximum sensitivi-ty must be used to measure heavy ions withlow flux. This poses no problem, as the limitervoltage requires 80 % to 90 % of the particleenergy/q, and the maximum output of theordinary onboard high-voltage power supplyis several keV.

Generally, the energy resolutionδE/E of aspherical electrostatic analyzer is approxi-mately proportional to (R2-R1)/(R1+R2),whereas the radii of the inner and outerspheres of the analyzer are R1 and R2, respec-tively. This means that if the size of the spheri-cal electrodes are increased while maintaininga constant separation between the radii of thetwo spheres, energy resolution will improveproportionally. Although limited satellite


resources will impose some restrictions in thisarea, if the center radius can be increased to100 mm, energy resolution would improve toapproximately 1/2. Similarly, the particle ener-gy/SV value may be expected to double, thusenabling measurement within a larger energyrange using the same high-voltage power sup-ply. Further, since resolution in the azimuthdirection is determined by the positioning res-olution of the analyzer, a larger center radiuswould provide higher angular resolution withthe same positional resolution. These pointsshould be taken into consideration whendesigning the actual ion measurement instru-ments.

6 Conclusion

Our calibration experiment with the proto-type analyzer offers a promising view into thepossibility of three-dimensional velocity dis-tribution measurement of plasma with a meas-uring instrument aboard a satellite featuring

three-axis control. In particular we have wit-nessed major advances in terms of solar windions and electron measurement in the contextof the L5 mission. We believe that the currentdesign will prove adequate to measure elec-trons in the solar wind. Future topics in thisarea include enlargement of the instrumentsize and the incorporation of identification ofion species via the TOF method. Finally, theanalyzer developed here will prove suitablefor use not only in the L5 mission but also inthe measurement of keV particles in theEarth’s magnetosphere.

Acknowledgement

We would like to express our sincerethanks to Mr. Junichi Komuro at the Scienceand Technology Information Group, Commu-nications Research Laboratory, for his cooper-ation in the design and manufacture of theprototype analyzer.


References1 G. Gloeckler, “Ion composition measurement techniques for space plasmas”, Rev. Sci. Instrum., 61, 3613,

1990.

2 K. A. Anderson, R. P. Lin, and H. D. Heetderks, “An experiment to measure interplanetary and solar elec-

trons”, IEEE Trans. Geosci. Instrumen., GE-16, 153, 1978.

3 L. A. Frank, D. M. Yeager, H. D. Owens, K. L. Ackerson, and M. L. English, “Quadrispherical LEPDEAS for

ISEE7s-1 and -2 plasma measurements”, IEEE Trans. Geosci. Electron., GE-16, 221, 1978.

4 S. J. Bame, J. R. Asbridge, H. E. Felthauser, J. P. Glore, G Paschmann, P. Hemmerich, K. Lehmann, and

H. Rosenbauer, “ISEE-1 and ISEE-2 fast plasma experiment and the ISEE-1 solar wind experiment”, IEEE

Trans. Geosci. Instrumen., GE-16, 216, 1978.

5 S. J. Bame, J. R. Asbridge, H. E. Felthauser, J. P. Glore, H. L. Hawk, and J. Chavez, “ISEE-C solar wind

plasma experiment”, IEEE Trans. Geosci. Instrumen., GE-16, 160, 1978.

6 G. L. Wrenn, J. F. E. Johnson, and J. J. Sojka, “The supra-thermal plasma analyzers on the ESA GEOS

satellites”, Space Sci. Instrum., 5, 271, 1981.

7 A. J. Coates, J. A. Bowles, R. A. Gowen, B. K. Hancock, A. D. Johnstone, and S. J. Kellock, “The AMPTE

UKS three dimentional ion experiment”, IEEE Trans. Geosci. Remote Sensing, GE-23, 287, 1985.

8 H. M. Shah, D. S. Hall, and C. P. Chaloner, “The electron experiment on AMPTE UKS”, IEEE Trans. Geosci.

Remote sensing, GE-23, 293, 1985.

9 A. D. Johnstone, S. J. kellock, A. J. Coates, M. F. Smith, T. Booker, and J. D. Winningham, “A space-borne

plasma analyzer for three dimensional measurements of the velocity distribution”, IEEE Trans. Nucl. Sci.,

NS-32, 139, 1985.

190 Journal of the National Institute of Information and Communications Technology Vol.51 Nos.1/2 2004

10 S. J. Bame, D. J. McComas, B. L. Barraclough, J. L. Phillips, K. J. Sofaly, J. C. Chavez, B. E. Goldstein, and

R. K. Sakurai, “The Ulysess solar wind plasma experiment”, Astron. Astrophys. Suppl. Sre., 92, 237, 1992.

11 L. A. Frank, K. L. Ackerson, J. A. Lee, M. R. English, and GT. L. Pickett, “The plasma instrumentation for the

Galileo mission”, Space Sci. Rev., 60, 283, 1992.

12 C. W. Carlson, D. W. Curtis, G. Paschmann, and W. Michael, “An instrument for rapidly measuring plasma

distribution functions with high resolution”, Adv. Space Sci., 2(7), 67, 1982.

13 M. J. Sablik, D. Golimowski, J. R. Sharber, and J. D. Winningham, “Computer simulation of a 360° field-of-

view ‘top-hat’ electrostatic analyzer”, Rev. Sci. Instrum., 59, 146, 1988.

14 T. Mukai, W. Miyake, T. Terasawa, and K. Hirao, “Observations of solar wind ions by the interplanetary

spacecraft Suisei (Planet-A)”, J. Geomag. Geoelectrc., Vol. 39, 377, 1987.

15 T. Mukai, S. Machida, Y. Saito, M. Hirahara, T. Terasawa, N. Kaya, T. Obara, M. Ejiri, and A. Nishida, “The

low energy particle (LEP) experiment onboard the GEOTAIL satellite”, J. Geomag. Geoelectrc., Vol. 46, 669,

1994.

16 S. Machida, Y. Saito, Y. Ito, and H. Hayakawa, “Instrument characteristics of the electron spectrum analyzer

(ESA) onboard the Planet-B mission and observational perspectives of the electron measurements”, Earth

Planets Space, Vol. 50, 207, 1998.

17 M. Akioka, K. Ohtaka, T. Nagatsuma, K. Marubashi, and W. Miyake, “L5 Mission and Observation of Inter-

planetary CME”, Journal of the Communications Research Laboratory, Vol. 49, No. 4, 13, 2002.

18 R. P. Lin, K. A. Anderson, S. Ashford, C. Carlson, D. Curtis, E. Ergun, D. Larson, J. McFadden, M. McCrthy,

G. K. Parks, H. Reme, J. M. Bosqued, J. Coutelier, F. Cotin, C. D’uston, K. P. Wenzen, T. R. Sanderson,

J. Henrion, and J. C. Ronnet, “A three-dimensional plasma and energetic particle investigation for the WIMD

spacecraft”, Space Sci. Rev., Vol. 71, 125, 1995.

19 W. Miyake and Y. Kazama, “Solar Wind Plasma Instrument for the L5 mission”, Journal of the Communica-

tions Research Laboratory, Vol. 49, No. 4, 27, 2002.

20 W. Miyake, E. Sagawa, A. Kawabe, and R. Yoshioka, “Calibration System for Charged Particle Analyzer”,

Review of the Communications Research Laboratory, Vol. 42, No. 2, 245, 1996.(in Japanese)

21 MCP assembly technical report, Hamamatsu Photonics Co., 1991.(in Japanese)

22 J. T. Gosling, J. R. Asbridge, S. J. Bame, and W. C. Feldman, “Effects of a long entrance aperture upon the

azimuthal response of spherical section electrostatic analyzers”, Rev. Sci. Instrum., 49, 1260, 1978.

23 T. Mukai and W. Miyake, “Transmission characteristics and fringing field effect of a 270° spherical electro-

static analyzer”, Rev. Sci. Instrum., 57, 49, 1986.

24 W. C. Feldman and E. Marsch, “Kinetic phenomena in the solar wind”, in ‘Cosmic winds and the helios-

phere’, 617, 1997.

25 R. P. Lin, “WIND observations of suprathermal electrons in the interplanetary medium”, Space Sci. Rev.,

Vol. 86, 61, 1998.

26 G. Borrini, J. T. Gosling, S. J. Bame, and W. C. Feldman, “Helium abundance enhancements in the solar

wind”, J. Geophys. Res., Vol. 87, 7370, 1982.

27 T. H. Zurbuchen, L. A. Fisk, N. A. Schwadron, and G. Gloeckler, “Observations of non-thermal properties of

heavy ions in the solar wind”, in ‘Acceleration and transport of energetic particles observed in the helios-

phere: ACE 2000 symposium’, 215, 2000.


MIYAKE Wataru, Dr. Sci.

Senior Researcher, Ionosphere andRadio Propagation Group, AppliedResearch and Standards Department

Space Weather

YAMAZAKI Atsushi, Dr. Sci.

Guest Researcher, Ionosphere andRadio Propagation Group, AppliedResearch and Standards Department

Earth and Planetary Science

192 Journal of the National Institute of Information and Communications Technology Vol.51 Nos.1/2 2004

calibration experiment of all-sky electrostatic analyzer...178 giott [9], ulysess [10], and galileo...

Documents