New practical method

for measurement of high-power LEDs

CORM 2008, Rensselaer Polytechnic Institute (RPI), Troy, NY, June 9-12, 2008

Yuqin Zong and Yoshi Ohno

National Institute of Standards and Technology

Gaithersburg, Maryland

Temperature Dependence of a High-power LED

0.70

0.75

0.80

0.85

0.90

0.95

1.00

1.05

1.10

5 15 25 35 45 55 65 75 85 95

Heat Sink Temperature (C)

Vo

lta

ge a

nd

Flu

x

no

rm

ali

zed

to

T=

25

C

5000

5500

6000

6500

7000

7500

8000

8500

9000

9500

10000

CC

T (

K)

Voltage (V)

Luminous Flux (lm)

CCT (K)

S1 at 350 mA dc forward current

Temperature issue in High-Power LED Measurements

However, SSL product manufacturers use LEDs typically

at much higher operating temperature.

Gaps to fill

LED manufacturers: Pulse measurement at Ta=Tj=25C

Lighting industry (traditional lamp manufacturers, Luminaire

manufacturers, testing laboratories):

DC measurement at Ta=25C.(LED thermal condition not defined)(Condition in LM-79, CIE 127 does not apply to high-power LEDs)

LEDs in ANY conditions (Tj of 10C ~ 100C) need

to be measured.

(Cold LED condition)

LED devices/packages

Existing Measurement Standards

CIE 127:2007

does not address issues on

high power LEDs

CIE TC2-46, TC2-58

CIE R2-36

We need a new, practical method

for high-power LED measurements

Simple Approach for High-Power LED Measurement

Use a metal-core PC board

Mount to a temperature-controlled heat sink (10 - 100 C).

Measure high power LEDs at a full DC current at thermal equilibrium,

or at short pulse.

- Heat sink temperature

- Pin temperature

- Case temperature

- Board temperature

.

Junction temperature (Tj) is the physical quantity of

an LED that reproduces exactly the LED operating

condition universally, anywhere.

Selection of the Temperature Reference

Not always available,

Not always reproducible,

Not always comparable,

Extra cost for the probe,

.

How to set, measure, and control the Tj?

Rated current

Set the heat sink

temp. to desired Tjand stabilize it.

Apply pulse

current, and

measure VF,j

VF,j

Optical quantities

measurement at Tj(in DC mode)

a few min. a few sec. a few min.

Feedback control

of heat sink

temperature

Tj VF,j

The New Practice Method

VF

VF,j

LED Heating Curves

0.95

0.96

0.97

0.98

0.99

1.00

1.01

1.02

0.000 0.010 0.020 0.030 0.040 0.050 0.060 0.070 0.080 0.090 0.100

Time (s)

Re

lati

ve

Va

lue

LED I

LED Vf

DET Sig

0.95

0.96

0.97

0.98

0.99

1.00

1.01

1.02

0 10 20 30 40 50 60

Time (s)

Re

lati

ve

Va

lue

LED I

LED Vf

DET Sig

0.95

0.96

0.97

0.98

0.99

1.00

1.01

1.02

0 10 20 30 40 50 60

Time (s)

Re

lati

ve

Va

lue

LED I

LED Vf

DET Sig

Mounted on a temperature-controller heat sink

0.95

0.96

0.97

0.98

0.99

1.00

1.01

1.02

0.000 0.010 0.020 0.030 0.040 0.050 0.060 0.070 0.080 0.090 0.100

Time (s)

Re

lati

ve

Va

lue

LED I

LED Vf

DET Sig

Un-mounted

Comparison of DC and Pulse Results under the same Vf

0.95

0.96

0.97

0.98

0.99

1.00

1.01

1.02

1.03

1.04

1.05

400 450 500 550 600 650 700

DC

/Pu

lse

Wavelength (nm)

Ratio of DC to pulse

Chromaticity x (DC-Pulse) -0.0001

Chromaticity y (DC-Pulse) 0.0001

Chromaticity u (DC-Pulse) 0.0001

Chromaticity v (DC-Pulse) 0.0002

CCT (K) (DC-Pulse) -3.1

Color Rendering Index (Ra) (DC-Pulse) 0.04

Luminous flux (DC/Pulse) 1.0002

Radiant flux (DC/Pulse) 0.9998

A phosphor-based

white LED:

If = 350 (mA),

Vf = 3.672 (V)

Pulse and DC measurements yield

the same results!!

Pulse: before thermal equilibrium.

DC: after thermal equilibrium.

Application of This Method

- IESNA, to develop a LM-xx Approved Method

for Photometric Measurement of High Power LEDs.

Develop a draft in ANSI SSL WG, then forward to the

SSL subcommittee in IESNA Testing Procedures

Committee.

- CIE

-

The whole measurement process can be fully automated.

Can also be used for measuring LED arrays.

Summary

LED can be measured at any junction temperature.

The reference condition is the LED junction temperature,

which makes mounting method and thermal resistance

between the LED and heat sink to be irrelevant to the result.

Pulses are used ONLY to determine the electrical forward

voltage. Optical quantities are measured (typically slow and

more difficult) in a normal DC operating condition.

We have developed a new, practical method for measurement

of high-power LEDs.

ACKNOWLEDGMENTS

Thank Dr. Cameron Miller, the DOE, Cree, Nichia,

and Seoul Semiconductor,