KNUT BREDEMEIERRALF POSCHMANNFRANZ SCHMIDT

For the measurement and evaluationof LEDs, there are well-defined rulesand instructions defining those mea-

suring procedures and instruments by me-ans of which important LED parameterscan be determined [1,2,3].The methods ofmeasurement described in the relevant li-terature regard the LED as a point sourceof light. For the development of modernluminaires, headlamps or units which useLEDs as light source, simulation tools suchas Speos or Asap are presently used. Indoing so, not only the LED as a point lightof source is described, but also the localdistribution of the LED radiation is takeninto consideration. The real radiation cha-racteristics of lamps Le (x, y, z, ϑ, ϕ) isdetermined either through complex mode-ling or through measurements. In manycases, ray data are used which have been

established by the lamp or LED manufac-turers and made available to the user.However, those data can also be measuredby a suitable measuring system, for exam-ple the goniophotometer type ›Rigo‹ ac-cording to Prof. Riemann [4,5,6,7]. Thesedevices use the image-resolved luminancemeasuring system for measuring ray datain the near field [8,9,10].

Measuring principle of thenear-field goniophotometer

The complete description of the radiationconditions of a body requires to know theradiance distribution Le (x, y, ϑ, ϕ, λ) inall surface points (x, y, z) ∈ of the body.These data can be obtained solely by ap-plying image-resolved measuring proce-

Near-Field GoniophotometerOPTICAL METROLOGY

Laser+Photonik 2 | 200720

Development of luminous objects withmeasured ray dataSPATIALLY RESOLVED RAYMEASUREMENTS ON LAMPS AND LEDS

The setting up of physical ray models is an extremely sophisticatedprocess, with those models being able to reflect reality in an onlyfinitely exact way. The measured ray data describe the light sourcewith corresponding accuracy. The data thus found can be exportedinto the data formats of all common simulation tools. Furthermore,they improve the simulation of any luminous sources and speed up

the development process.

1 The picture coordinates and camera positions can be converted in spacial directions

Object

ϕ

ϑ

fCCD-Sensor

(x',y')

Ray calculation and goniometer coordinate system

© Laser+Photonik

© 2

007

Car

l Han

ser

Ver

lag,

Mun

ich,

Ger

man

y

ww

w.la

ser-

phot

onic

s.eu

N

ot fo

r us

e in

inte

rnet

or

intr

anet

site

s. N

ot fo

r el

ectr

onic

dis

trib

utio

n.

dures. In the case of the near-field gonio-photometer type ›Rigo801‹, a measuringcamera is moved by the mechanical sy-stem of the goniophotometer on a spheri-cal surface around the object of measure-ment. While doing so, the measuring ca-mera captures image data in awell-defined angular grid, with these databeing used to calculate a ray data recordL (x, y, ϑ, ϕ). Usually, the measuring ca-mera is equipped with a specially adapted

V (λ)-filter to measure luminances. Byusing other suitable filters, for examplecolor-matching functions can be realizedto obtain multichannel color informationby running several series of measurement.All luminance images taken together re-sult in a four-dimensional data field L (x’,y’, ϑK, ϕK). Through the optical image per-formed by the camera lens, the imagecoordinates (x’, y’) – together with the in-formation about the camera positions

Near-field Goniophotometer

2 | 2007

V

2 The far-field luminous intensity distribution can be calculated from the ray data

Radiation characteristic I (ϑ, ϕ)

3 The conversion programm calculates the intersection points of the rays with the evelop

Recording the ray data on a cylindrical enveloping geometry

© 2

007

Car

l Han

ser

Ver

lag,

Mun

ich,

Ger

man

y

ww

w.la

ser-

phot

onic

s.eu

N

ot fo

r us

e in

inte

rnet

or

intr

anet

site

s. N

ot fo

r el

ectr

onic

dis

trib

utio

n.

( ϑK, ϕK) - can be converted into spatialdirections (Figure 1). The data field mea-sured contains all photometric informa-tion about the light radiation. From theray data, a number of derived measuredquantities such as the measured luminousflux Φ and the far-field luminous intensitydistribution I (ϑ, ϕ) can be calculated(Figure 2). A conversion program allowsthe ray data to be output in the desireddata format. Here, a selectable number of

rays can be exported from the TechnoTeamformat into the standard formats for Asap,Speos, LightTools, Zemax and LucidShape.As the real geometry of the object of me-asurement cannot be determined by the

measuring system, the starting points ofthe rays are found at first on the spheri-cal surface described by the entrance pu-pil of the lens. The conversion programtransforms these starting coordinatesonto a selectable enveloping geometry bycalculating the intersection points of therays with this envelope (Figure 3). Rayswhich do not intersect the envelope, willnot be exported into the target format.The enveloping shapes presently availableare a circle, a cylinder, and a cuboid. Thegeometrical parameters are usually chosensuch that an optimum convex envelope

around the object of measurement resultsin order to avoid any undercuts with pos-sibly existing nearby objects during simu-lation. The devices of the ›RiGO801‹-seriesallow the measurement of objects of most

Near-field GoniophotometerOPTICAL METROLOGY

Laser+Photonik 2 | 200722

4 Example of the strong distortion on glass elements of a D2 lamp

5 For white LEDs it isnecessary to knowthe dependences onthe wavelengthbecause they areposition-dependent

3800

Spec

tral

sig

nal d

ensi

ty

0.0005

0.0010

0.0015

0.0020

0.0025

0.0030

0.0035

0.0040

430 480 530 580 630 680 730 780

Wavelength

Spectrum of a white LED

6 Color image of an LED cluster

7 In the middle, the radiation characteristic of the LED (on the left) is shown, and the simulated optical path of the bicycle headlamp on the right

Development of an LED bicycle headlamp

© Laser+Photonik

© 2

007

Car

l Han

ser

Ver

lag,

Mun

ich,

Ger

man

y

ww

w.la

ser-

phot

onic

s.eu

N

ot fo

r us

e in

inte

rnet

or

intr

anet

site

s. N

ot fo

r el

ectr

onic

dis

trib

utio

n.

various sizes, ranging from LEDs to lampsand luminaires having a length of2000 mm. To be able to measure the raycharacteristic of small objects such asLEDs or small lamps (up to 50 x 50 x

50 mm3), a special LED goniophotometerwas designed (Title picture). Using thesedevices, only those objects can be measu-red which either fit into the goniometercollision-free, or can completely be cove-red by the imaging system. On the otherhand, the objects must not be too smallas otherwise the mechanical tolerances aswell as the imaging characteristics cannotbe measured with sufficient accuracy.

Ray data of LEDs and arc lamps

In order to obtain realistic simulation re-sults in the development of optical com-ponents, the radiation characteristics ofthe light sources must be described veryexactly. For this, various models of diffe-

ring complexity are availa-ble ranging from a simpledistribution of direction ofa point source of light andcomplex physical models toreally measured ray data di-stributions. In many appli-cations, either a physicalmodel or a model based onmeasured ray data is neces-sary because optical com-ponents are arranged in theimmediate vicinity of thelamp geometry.

The setting up of physi-cal models is often extre-mely complex and requiresan exact definition of thegeometries and of both theoptical and thermal proper-ties of all optically relevantelements. Particularly in thecase of modern LED techno-logies and arc lamps, thework to be invested in rea-

listic models is immense. As an example,Figure 4 shows the optical distortions ona D2 lamp, which are difficult to model.

Measured ray data describe the reallight source, thus making a complex phy-sical model superfluous. The combinationof the measured data with geometrical in-formation about the geometry of the lightsource makes it possible to take the in-fluence of the light source in the opticalpath (vignetting) into account. The realray data distributions also consider themanufacturing tolerances, which have adisturbing effect particularly in the caseof small lamps such as LEDs. Furthermore,several samples of a light source can bemeasured to determine the variability ofthe real parameters. Another importantaspect is the determination of the spec-tral properties of light sources. The com-plete spatially resolved measurement ofthe spectral distribution is metrologicallynot practicable.

In the case of color LEDs (narrowband),the integral spectral distribution is suffi-cient in most cases to carry out simula-tions. For white LEDs, however, these dataare not enough. It is necessary to knowthe dependences on the wavelength (Fi-gure 5) particularly because they are po-sition-dependent (blue diode and yellowdye geometrically separate in Figure 6),i.e., they have a five-dimensional charac-teristic L (x, y, ϑ, ϕ, λ). Some first

Near-field Goniophotometer OPTICAL METROLOGY

2 | 2007

C O N TA C T

TechnoTeam Bildverarbeitung GmbH,98693 Ilmenau, Phone +49 (0) 36 77 /46 24 -0,Fax +49 (0) 36 77 /46 24 -10,www.technoteam.de

8 In the case of the control element in the passenger compart-ment (on the left), the LED radiation is injected via optical wave-guide (in the middle); on the right: simulated ray distribution

V

© 2

007

Car

l Han

ser

Ver

lag,

Mun

ich,

Ger

man

y

ww

w.la

ser-

phot

onic

s.eu

N

ot fo

r us

e in

inte

rnet

or

intr

anet

site

s. N

ot fo

r el

ectr

onic

dis

trib

utio

n.

useable measuring results were obtainedwith color filters. By inserting a blue anda yellow filter, the ray data could be se-parated into blue and yellow rays in twomeasurements. Furthermore, the direc-tion-resolved measurement of spectra bymeans of a spectrometer (option) is pos-sible. The spectrometer is moved – toget-her with the camera – by the mechanicalsystem of the goniophotometer on a sphe-rical surface. The development of an LEDbicycle headlamp serves as an example(Figure 7). Here, both the reflector andthe cap were optimized by using measu-red ray data of an LED (side emitting). Inthis way, the system was realized withonly two iterations. Applications in thecase of which optical devices are arrangedin the extreme near field can practicallyhardly ever be simulated without the realray data. An example is the light injectioninto optical waveguides by means of anLED. Figure 8 shows a control element forthe passenger compartment developed inthis way. The near-field goniophotometerallows not only the complete ray data tobe obtained, but also single high-resolu-tion luminance images and measuring se-ries to be taken with the ›LMK2000‹ lumi-nance measuring software. ■

Summary: High-speed development process

The Rigo goniophotometers accordingto Prof. Riemann utilize the image-re-solved luminance measurement techni-que for measuring ray data in the nearfield. Besides, this method permits thespatially resolved measurement withsuitable color filters and, combinedwith a spectrometer, also the direction-resolved measurement of the spectralcharacteristics. The ray data thus deter-mined can be exported, for example,into the data formats Asap, Speos,Light-Tools, Zemax and LucidShape. Thedevelopment process is considerablyspeeded up even in the case of the si-mulation of a light injection into an op-tical waveguide by means of LEDs.

LITERATURE [1] CIE: ›Measurement of LEDs‹; Publication CIE127-

1997, 1997

[2] Muray, K.; Schanda, G.; Schanda, J.: ›Filter fabrica-

tion techniques for LED photometry‹; CIE LED Sympo-

sium 97 on Standard Methods for Specifying and Meas-

uring LED Characteristics, CIEx013, 1997, S. 61-64

[3] Jones, C.F.: ›Colorimetry, chromaticity space and

LEDs‹; SPIE Conference on Illumination and Source En-

gineering, San Diego, California, July 1998, SPIE Vol.

3428, S. 100-114

[4] Poschmann, R.; Riemann, M.; Schmidt, F.: ›Verfah-

ren und Anordnung zur Messung der Lichtstärkevertei-

lung von Leuchten und Lampen‹; Patent DE 41 10 574

v. 30.03.1991

[5] Riemann, M.; Schmidt, F.; Poschmann, R.: ›Zur Be-

stimmung der Lichtstärkeverteilung von Leuchten inner-

halb der fotometrischen Grenzentfernung mittels bild-

auflösendem Gonimeter‹ Licht 7-8/1993; S. 592 – 597

[6] ›Anordnung zur Messung der Lichtstärkeverteilung

von Leuchten und Lampen‹; TechnoTeam Bildverarbei-

tung, Gebrauchsmuster 297 06 488.6 v. 11.04.1997

[7] Nolte, R.; Bredemeier, K.; Poschmann, R.; Schmidt,

F.: ›10 Jahre Nahfeldgoniophotometer – Grenzen und

Möglichkeiten‹; Tagung Licht 2004, Dortmund

[8] Krüger, U.: ›Ausgewählte Aspekte der ortsaufgelös-

ten Erfassung stationärer Licht- und Farbverteilungen‹;

Dissertation TU Ilmenau, Fakultät für Elektrotechnik

und Informationstechnik, 2004

[9] Bredemeier, K.; Poschmann, R.; Schmidt, F.; ›Nah-

feldgoniophotometer – Systeme zur Messung der Licht-

verteilung an Leuchten, Lampen und LED‹; Licht 2006,

Bern 10.-13.09. 2006, Tagungsband (CD)

[10] Krüger, U.; Schmidt, F.; Wolf, St.: ›Applications of

image resolved light and colour measurement‹; Light &

Engineering 3/2005, Znack Publishing House, Moscow

AUTHORSDipl-Ing. Knut BREDEMEIER and Dr.-Ing. Ralf POSCH-

MANN are development engineers with TechnoTeam

Bildverarbeitung in Ilmenau, Germany.

Dr.-Ing. habil. Franz SCHMIDT is CEO of the TechnoTeam

Bildverarbeitung.

Near-field GoniophotometerOPTICAL METROLOGY

Laser+Photonik 2 | 200724© 2

007

Car

l Han

ser

Ver

lag,

Mun

ich,

Ger

man

y

ww

w.la

ser-

phot

onic

s.eu

N

ot fo

r us

e in

inte

rnet

or

intr

anet

site

s. N

ot fo

r el

ectr

onic

dis

trib

utio

n.