TG 1103

FEBRUARY 1970

Copy No. -.

'Technical Memorandum

SOLAR PANEL TEST SET

by W. E. RAY

"JUNl 93 !9?

THE JOHNS HOPKINS UNIVERSITY a APPLIED P'YSICS LABORATORY

Th w do-umont hos been opproved for publicreleawe and sole its distribution nl unlimited

Fii.,IUAR 1970

KI Technical Merforandum

SOLAR PANEL TEST SEThi'I by W. E. RAY

TH ON OKN UIEST1PLE PYISLBRTR8621Georia venu, SlverSprng, aryand 091Oprtn ne otatN w6-64- ihteDprmn fteNv

8621s Geogi Avene, St htiuini nilvterdpig.arln 01

I

APPLIED PHYSICS LASORATORY

1tVG* OPM" UMAM~ABSTRACT

IThis report describes the Solar Panel Test Set de-

veloped for testing solar cell panels in artificial sunlightrat an equivalent sunlight intensity of 140 mW/cm 2 . The

test set uses iodine-quartz (tungsten) lamps as the radiant-energy source, and the emerging radiation is uniformly re-flected and totally diffused. An air conditioner, which ispart of the test set, provides the cooling air necessary tocontrol the temperature of the solar panel under test. Themethods of calibrating the test set are described, and theaccuracy of the measurement obtained when using artificiallight as the radiation source is discussed.

[

II

I

l~g JOMMI HOPKINS WWRSTV

7 APPLIED PHYSICS LABORATORY

CONTENTS

List of Illustrations • • vii

- 1. Introduction . . . . 12. Description . . . . 3

.. 3. Theory of Operation . • • 7

4. Calibration . . . 21

5. Performance . . . . 23

i

S.v

-b

.Is

II

I

KAPPLIED PHYSICS LABORATORY

ILVMm e A*Y.AM"M

IILLUSTRATIONSI

1 Solar Panel Test Set 4

2 Sun Spectral Distribution, Solar CellResponse versus Wavelength . .8

3 Iodine-Quartz Distribution, Solar CellResponse versus Wavelength . . 10I

4 Illuminator Emerging-Intensity Pattern. 12

5 Directional Reflectance versus Viewing[ Angle, Source at -70 . . . 13

6 Iodine-Quartz Lamp Color Temperature[as a Function of Supply Voltage . 14

7 Iodine-Quartz Lamp Output in Lumens as aFunction of Supply Voltage .15

8 Solar Panel Test Fixture 0 17

9 Power Control System . . 1910 Percent Variance of Irradiation Intensity in

Test Plane Measured by 2 x 2 cm,10 ohm-cm Solar Cell; 1sc 140 mA . 22

11 Typical Current-Voltage Curve . 24

I i

I - vii -

APPLIED PHYSICS LABORATORY

I1. INTRODUCTIONI

I The Solar Panel Test Set is used to functionallytest solar panels of current and future APL-built satellites,using controlled illumination to test the solar panel poweroutput in accordance with the design specifications. Inaddition, the test set is helpful in determining and isolating

I malfunctions.

The sunlight method for testing solar panels hasalways presented a problem. There are few cloud-freedays with uniform light intensity; therefore, the panelsmust be tested with existing sunlight conditions, and thenthe resulting data must be normalized. In the past, many

-b tests on the solar panels and the satellite were delayed ornot run because of insufficient sunlight at the scheduledtest time. Also, control of the temperature of solar panelsbeing tested in sunlight is difficult, if not impossible,whereas the artificial sunlight of the test set permits thetemperature of the solar panel to be measured and con-trolled.

II'II

I -1 -

I

!TlM jaw*s eoINiS VuVWiVugT

APPLIED PHYSICS LABORATORY

2. DESCRIPTIONI

3 The design of the test set permits solar panels aslarge as 20 by 55 inches to be tested, with the illuminationintensity on the surface of the panel being held within 2%of an equivalent sunlight air mass zero (AM = 0) intensityof 140 mW/cm 2 . Larger panels may be tested, but thereis a slight degradation of uniformity in the illuminationplane. The test set is illustrated in Fig. 1. The majorassemblies are the illuminator, solar panel test fixture,air conditioner, power control system, instrumentationrack, and calibration scanner.

The illuminator consists of a rectangular box, ap-proximately 95 inches long, 72 inches wide, and 72 incheshigh. The sides and roof of the box are made of aluminumpanels attached to an aluminum structure that is supportedby four legs. A radiant-energy source, located on theceiling of the illuminator, consists of twenty 1000-watt,2000-hour iodine-quartz (tungsten) lamps that are capableof maintaining a constant intensity output throughout theirrated life. The lamps, mounted in ceramic lamp socketsKthat are supported on bracket assemblies, are suspendedapproximately 7 inches from the ceiling of the illuminator.The top portion of the illuminator consists of a cold-airplenum and a hot-air exhaust system. The pressurizedair in the cold-air plenum is used to cool the lamp sockets,while the hot-air exhaust system pulls the hot air generatedby the lamps out through the top of the illuminator and intothe air conditioner.

The solar panel test fixture is a movable carriagethat supports the solar panel being tested and allows it tobe located in a predetermined position beneath the radiant-energy source. A reversible motor, mounted on the baseof the test fixture and coupled to a counter, permits thevertical position of the test fixture to be set accurately.

K-3-

K PR~pgD!tIG PAGE BLANK

APPLIED PHYSICS LABORATORYG*LVM SPgIUS MANYLAN

POWERCONTOL SSTEM COLD-AIR SUPPLY DUCT ~ . AIR CONDITIONER

NOT-AIR RETURN DUCT H0:::::,

iNOT-AIR RETURN PLENUM

'~SECONDFLOOR LEVE

SOLAR PANEL UNDER TEST,

OR CALIBRATION SCANNER

INSTRUMENTATION

RACK 1Fig. I SOLAR PANEL TEST SET I

4-

THE Joe 4OP IU UIVeIIYAPPLIED PHYSICS LABORATORY

'B I wS UIWme M*A.N

An airflow and cooling system for the solar panelunder test is provided by a 7j-ton air conditioner. Air isfed into the plenum chamber through regulators that directthe air to the underside of the solar panel. The tempera-ture of the solar panel is controlled by adjusting the direc-tion of the air that passes through the regulators.

The power control system includes a power rackand a power distribution panel. Five voltage regulatorsfor the lamp circuits are located in the power rack, alongwith the necessary cooling fans. The power distributionpanel contains the relays and interlocks that operate thelamps and the air conditioner. The interlocking relaysare so placed in the system that an air-conditioner failureresulting in either no circulation of air or improper cool-ing of the returned hot air would cause the power to the[ lamp voltage regulator to be turned off.

The instrumentation rack contains the equipmentrequired for measuring the output current and voltage of[ the solar panel under test; these measurements are plottedon an X-Y recorder. The temperature-measuring instru-ments, the load, and the digital voltmeter are also locatedin the instrumentation rack.

The calibration scanner is a device that is attached[I to the solar panel test fixture whenever calibration is tobe performed. By means of a standard solar cell andthermistor, it measures the intensity at each point in thesolar-panel plane. The standard cell must be of the samedimensions (2 by 2 cm) and spectral characteristics asthose of the solar panel. The standard cell and the therm-istor are mounted on a platform that can be remotely ma-neuvered to any location within the illuminator. The stan-dard cell measures the actual intensity, and the thermis-tor permits the intensity measurements to be correlatedwith temperature.

[[

~-5-

THE .JON" MMPINS IUdNUIY

APPLIED PHYSICS LABORATORYUILVERM SpUnig MARYLAO

I3. THEORY OF OPERATION

I The radiation emerging from the illuminator (inwhich the radiation from each lamp is uniformly reflectedand totally diffused) is independent of lamp position and$1 differences in lamp intensities. The proper intensity ofillumination is achieved by maintaining a minimum numberof lamps and by providing diffuse reflectors with high coef-ficients of diffusion and reflectivity. A white paint (3MVelvet White) is used on the walls as the diffuse-reflectormaterial. Specular reflectors are used on the lower wallsof the illuminator, so that a uniform intensity pattern isobtained on the surface of the solar panel, The intensity

I of emerging illumination is held uniform to within 2%across the surface of the solar panel.

In its operation the Solar Panel Test Set does notsimulate the spectral distribution of the sun at air masszero. The simulation of the sun's spectral distribution

T is a science in itself, and an attempt at such a simulationin near space would needlessly complicate the test set.

The electrical output of a solar cell, Isc (short-circuit current), is a linear function of the intensity of theincident solar radiation. Figure 2 shows the spectral dis-tribution of the sun and the spectral response of a siliconsolar cell. Both are normalized for illustration purposes.The short-circuit current of the solar cell is expressed byJ the equation:

I sc = IKRac() - Rs(A)dX, (1)0

i where

I -short-circuit current of solar cell,Sc

1 -7-

PRECEING PASE BLANK

THE J OMNS HOPKINS UNIVERSITY

APPLIED PHYSICS LABORATORYIiuvt* UWo"14 MASYLANO

zz

C L

E

w I W

4 w 4

L V)0

00C

on

C14

F--

Th[ JOHN$ MOPKINS UNIVENITY

APPLIED PHYSICS LABORATORYSILVEW SMiG MAROY.AND

!K = scale factor to correct relative response

to absolute response,

Rsc relative spectral response of solar cell,

I s = relative spectral distribution of the sun.

As Eq. (1) shows, the solar cell does not convert allradiant energy of the sun into electrical energy but onlyconverts that energy falling within the spectral responseof the solar cell.

Now, consider the spectral distribution of theiodine-quartz lamps and the spectral response of tihe so-arcell, as shown in Fig. 3. The short-circuit currbnt inthis case is expressed by the equation:

Ic KR(X) R (X)dX , (2) f

where

R relative spectral distribution of tungoten.T

Similarly, the solar cell will convert that portion of theradiant energy of the iodine-quartz lamps that falls withinthe spectr~l response of the solar cell.

Although the spectral distributions of tne iodine-quartz lamps and the sun are different, the solar cell is

k responsive to a certain portion of the spectral distributionof each. By adjusting the number of lamps used and thelamp color temperature, the electrical output of the solarcell exposed to the radiant energy of the test set can bemade to produce the same electrical output that it would

produce in space. The test set intensity is controlled bya standard solar cell, which has been carefully calibratedin sunlight and also in a solar simulator. It is importantthat the solar cells or solar arrays being tested have thesame spectral response as the standard solar cell; there-fore, they are all selected from the same production lot.

I -9-

TmE jowNO MOEINS UmivIETyvAPPLIED PHYSICS LABORATORY

SA-VtO GM MGa MARYLAND0

z

0.

I-x

0 z

-0

x Ix

X

z

'4

zz

w m

zU

10 I-

ITHE .OHN$ HOPKINS UNIVEMITY

APPLIED PHYSICS LABORATORY$SOLV90 OMP tP. MAOYLANO

IAs can be seen in Fig. 3, the electrical output of the solarcell will change if there is a change in the spectral response.

The illuminator is so designed and built that theradiation energy reaching the array under test is totallydiffused from a multiunit radiant-energy source ratherthan collimated from a single-point radiant-energy source.The total diffusion of the light is accomplished by distribu-ting the iodine-quai - lamps on the ceiling of the illumina-tor and by painting t. e ceiling and wall of the illuminatorwith a white paint that has a high coefficient of diffusionand reflectivity. Specular reflectors are placed on thelower walls of the illuminator so that the intensity at thebottom of the test plane will not vary more than 2% acrossthe area where the array will be located. Figure 4 showsthe irradiation at the test plane with and without the re-flectors at the bottom of the illuminator. Figure 5 showsthe directional reflectance of the 3M Velvet White paim,used on the walls of the illuminator, at various viewing

T angles when illuminated from an angle of -700. The paint '&does not burnish to a higher gloss, and thus the walls of

the illuminator can be kept clear by periodic cleaning witha mild detergent.

The characteristics of the iodine-quartz lamp areshown in Figs. 6 and 7. Figure 6 shows the color tem-perature as a function of the supply voltage. As can beseen from the graph, a 3% change in voltage from the

I power supply will cause a change of color temperature ofless than 1%. Figure 7 shows the lamp output in lumensas a function of the supply voltage.

The AC voltage across the lamps is regulated byfive voltage regulators (four lamps for each regulator).

I The regulators are capable of maintaining the load voltageto within ±1%, with input voltage variations of ±10%.

SI The lamps and the illuminator are protectedagainst excessive temperatures by means of interlockswith the plenum airflow and temperature sensors. Loss

1 -11 -

I

ThEJOHs HPMINS UNIVERSIYAPPLIED PHYSICS LABORATORY

I&Vv[O SM%§., IN14YLAND

w

u Ixz wI

xI z

u -

12 -

THE JON 'MO$KIN UNIVERSITYAPPLIED PHYSICS LABORATORY

i IUtS gim w MA&~rYANO

I

12

,[ ,, 11

,,IL 10 -SOURCE _ _ _ -.

z

I-49_ _4

0u

I~j "---,"0 3M VELVET COATING 202.A10"

I LI 5 3A2

S0 10 20 30 40 50 60 70 soVIEWING ANGLE, V (degrees)

Fig. 5 DIRECTIONAL REFLECTANCE VERSUS VIEWING ANGLE,SOURCE AT -70I A

II

1 -13-

I

mla JOH. IOPN* UNI~VgarnrAPPLIED PHYSICS LABORATORY

...v°U S9.... i 7',U

'09

,0 _ _ _ _ _ _ _ _ _106

"103-- /

0

Jo 97ua

I-4 94

zuwu 91-_

88

40 60 80 100 120 140 160 180 200

VOLTAGE

Fig. 6 IODINE-QUARTZ LAMP COLOR TEMPERATURE AS AFUNCTION OF SUPPLY VOLTAGE

-14 -

"m 4N NOPqKNA UoilWAM'I~'

APIIED PHYSICS LADORATOfRY

II

1 175 _.

150 - -..

z

wS125

A loo

W 7510

I-I

25

j 040 60 s0 100 120 140 160 180 200

I VOLTAGE

Fig. 7 IODINE-QUARTZ LAMP OUTPUT IN LUMENSAS A FUNCTION OF SUPPLY VOLTAGE

i -15 -

K

APPLIED PHYSICS LABORATORY

of the cold airflows will open an air-vane switch locatedin the outlet duct, and too high temperature of the air re-turned to the air conditioner will open the temperature-sensor switch. The actuation of either switch will deen-ergize the lamps.

The solar paneltest fixture (Fig. 8) provides ameans for positioning the solar panel vertically under theilluminator, with a high degree of repeatability, by meansof a factional-horsepower gear motor that is mechanicallycoupled to the four legs of the fixture. A mechanicalcounter attached to the test fixture and geared to the drivemotor indicates the height of the test fixture to the nearest10 mils (0. 010 inch). The total possible vertical travelof the solar panel mounted in the test fixture is approxi-mately 18 inches.

The test fixture also provides the controlled coolingneeded for accurate measurement of the solar panel output.The temperature across the panel is maintained constantby means of the airflow to the panel through the air ductsystem,which is an integral part of the solar panel test fix-ture. The top of the air plenum is fitted with four airregisters, and the airflow through these registers is con-trolled by two sets of deflection vanes set in the registerat 900 angles to each other. In addition, each air registerhas an airflow damper that controls the air passing throughit.

The solar panel bracket, which holds the solarpanel being tested on the test fixture, is designed to ac-commodate the Navy Navigation Satellite, OSCAR, solarpanels. The bracket prevents any movement of the bladethat could be caused by the cooling air from the air regis-ter putting pressure on the underside of the solar panel.A blade-mounting adapter can be attached to accomwodaiesolar arrays of different shapes, such as those used onthe GEOS-type satellites. The bracket is also designedto accommodate the calibration scanner.

- 16 -

APPLIiD PHYSICS LABORATORY

AIR REGISTERS WITH

TDEFLECTION VANES dm

KSOLAR PANEL

HOLD- DOWN-dr -ARPEU

CARRIAGE

IFig. 8 SOLAR PANEL TEST FIXTURE

'17

THInt £HW MPKINS UNI33STYAPPLIED PHYSICS LABORATORY

The equipment located in the instrumentation rackmeasures the electrical output characteristics of the solararray under test and, in most cases, presents the infor-mation in the form of a permanent current-voltage (I-V)curve plotted on the X-Y recorder. The voltage potentialacross a circuit of an array under test is plotted on theX-axis of the recorder. The current is recorded on theY-axis by means of the voltage drop across a precisionresistor. The temperature of the solar panel is monitoredby thermistor probes that are attached to the solar panelduring test. The digital thermometer indicates the tem-perature in degrees centigrade from any one of four se-lectable channels. The digital voltmeter and the precisionammeter are used to calibrate the system.

The air-conditioning system is a closed-loop sys-tem; that is, the hot air generated by the iodine-quartzlamps and the solar panel is returned to the air conditionerto be cooled and recirculated. The unit has a two-stagecompressor. The cool air is fed into the inner air plenumat the top of the illuminator for cooling the lamps and lampsockets and is also fed to the solar panel test fixture. Thecool-air duct is connected to the solar panel test fixtureby means of a flexible coupling, which permits the testfixture to be raised and lowered while the air conditioneris operating.

The electrical distribution system and the safetyinterlock system are shown in Fig. 9.

- 18 -

Twe jO94Ns MOftsis U*Iv3UaITYAPPLIED PHYSICS LABORATORY

- So*LV9M gp~osg MARYLAND

0

UJI z

00- IL

. : w

-on II

un0

0u

01 00 J

ILI 4 .- -

tI I-

-uII i; o

__ 41Ii _ _ _ _ _ _

T A44M00 HOPKINS UNIVI[NftIT'

APPLIED PHYSICS LABORATORY5.v~S UP lr. M*SI.tANO

4. CALIBRATION

The illuminator is calibrated to determine the uni-formity of the intensity level in a plane at the test level.The alibration scanner consists of a two-axis traversingmechanism to which a platform is attached in such a waythat the platform can travel to any location within the il-luminator at a predetermined height. Mounted on the plat-form are a standard solar cell and a thermistor with leadsgoing from the platform to the instrumentation rack. Thestandard solar cell short-circuit current is plotted on theY-axis of the X-Y recorder, and the location of the plat-form along the long axis of the illuminator is plotted onthe X-axis. By moving the platform in 2-inch increments

along the short axis and traversing the platform along thelong axis, the short-circuit current of the standard solarcell can be plotted for the entire test-plane area. Sinceboth the current and voliage characteristics of a solar celltemperature-dependent, the temperature of the standardcell is concurrently displayed on the digital thermometer.

Figure 10 shows the percent variance of the irra-diation intensity in the test plane as measured by a 2 by2 cm, 10 ohm-cm solar cell.

- 21

' LANK

THE JOHNS H4OPKINS UNIVERSITY

APPI IED PHYSICS LABORATORYSIL~tf tPft.PG 6MA*VLAND

95 INCHES -

8+

-40;

Fig. 10PERCENTOARIANCE OF OSCAAI R INESITYR IN8ES- PANMEASREDBY x2 c, 1 oh-cm OLA CEL; 'c 10 m

- 22 -

APPLIED PHYSICS LABORATORY

5. PERFORMANCE

Usually, the solar array manufacturer supplies anI-V curve for each circuit on each array delivered to APL.These curves are produced from data obtained by meansof the manufacturer's solar simulator tests or from theJet Propulsion Laboratory's Table Mountain, California,facility tests. These test data are corrected to reflectthe output of the circuits in outer space (AM = 0) and at anoperating temperature of 280C.

The I-V curves obtained from the APL Solar PanelTest Set are compared with the I-V curves supplied by themanufacturer to determine if the purchase specification ismet. In addition, I-V curves are obtained from the APLSolar Pane1 Test Set after the arrays have completed suchqualification and evaluation tests as vibration and thermalvacuum tests. These I-V curves are then compared withthe I-V curves initially obtained to deterrm~ine if the testscaused any degradation in the output of any circuit on thee.rrays.

Figure 11 shows a typical I-V curve obtained fromthe APL Solar Panel Test Set compared with the I-V curvesupplied by the manufacturer for the same circuit. Bothcurves have been corrected for an AM = 0 output at 28*C.In general, the curves from APL and from the manufacturerare in good agreement, with the greatest deviation occurringat open-circuit voltage (Voc). Typically, the short-circuitcurrent measurements agree to within 2% and the open-circuit voltage to within 4%. The greater deviation in theopen-circuit voltage measurement is due to its being moresensitive to temperature than is the short-circuit current.The packing density of the solar cells on the substrate doesnot permit a thermocouple or a sensistor to be placed sothat a reliable temperature measurement can be obtained.Nevertheless, the APL Solar Panel Test Set has beenproven to be a reliabt instrur ,.nt for the evaluation ofsolar cell arrays.

- 23 -

THE JO44NI HOPKINS UPIVRSIrY

APPLIED PHYSICS LABORATORYSILV~tO UIIWE MARYLAND

400 .h F 1...IMANUFACTURER'*S DATA

350

300 -_____ APL SOLAR PANELI

E

zw 200

31 SOLAR CELLS IN SEIRIES)

150 3 SOLAR CELLS IN PARALLEL2 okm-cm N on P SOLAR CELLS

100

so -~

VOLTAGE

Fig. 11 TYPICAL CURRENT-VOLTAGE CURVE

-24-

_ __L Ic LASSIFIEDS. ir l ~ a%-uzcation

The .ohns iHopkins Uiniver sity Applied Physics Lab. I~

Solar Panel Test Set

4 ')SC RIP 7IVE NOTES ('pe of report and inclusive dates)I

AUI THORtI (First name, middle initial, Iast nitme)

William E. Ray4

45 REPORT DATE 7a. TOTAL NO OF PAGES 7b. .1O OF REFS

February 1970 2ft. CONTRACT OR GRANT NO 90. ORIGINATOR'S REPORT NUMBER($)

NOw 62 0604-ch,. PROJECT NOTG10

9b. OTHER REPORT NO(SI (Any other numbers that may be assignedthis report)

d.

IC DISTRIBUTION STATEWrNT

This report has been approved for public release and sale; itsdistribution is unlimited.

It SUPPLEMENTAiFY NOTES 12 SPONSORING MILITARY ACTIVITY

Strategic Systems Project Office

13 ABSTRACT

T4"i-report describes the Solar Panel Test Set developed for testingsolar icell panels in artificial sunlight at an equivalent sunlig!'t intensity of 140mW/c~yt2 . The test set uses iodine-quartz (tungsten) lamps as the radiant-energysoirce, and the emerging radiation is uniformly reflected and totally diffused. Anai' condr Loner , which is part of the test set, provides the cooling air necessary to

f-onti-ol thil temporature of the solar panel under test. The methods of calibratingthe test set ars? described, and the accuracy of the measurements obtained wheIlnusing artificial light as the radiation source is discussed.

DDFNOV,1473 'C ASTI)Securitv ClassifIcation

111SUASSIFTEDSecurity Classification

14.KEY *OROS

IlluminatorIllumination planeDiffused multiunit radiant -energy sourceCurrent-voltage curvesSolar cellSolar panelSolar arrayStandard solar cell

UNCLASSIFIEL)

Security Classification