1
Junction Temperature Characterization of High Power Light Emitting Diodes
Moon-Hwan Chang, Diganta Das, and Michael Pecht*
Center for Advanced Life Cycle Engineering
Department of Mechanical Engineering
University of Maryland
College Park, MD 20742
* Corresponding Author
Abstract
High junction temperature affects the optical performance and reliability of high power LEDs. This paper
presents a study of the results of junction temperature measurement tests under various drive currents and ambient
temperatures carried out on 3W white InGaN high power LEDs. The goal was to improve the accuracy of junction
temperature characterization of the LEDs under test. Test results showed that 200 steps of pulse with 0.061% duty
cycle can be used practically to construct reference plot of junction temperature (Tj) vs. forward voltage (Vf) at
specific current levels for 3W high power LEDs so that Joule heating of LEDs can be ignored. High power LEDs
mounted on an aluminum metal printed circuit board were tested inside a temperature chamber operated in
temperature steps to reach a thermal equilibrium condition between the chamber and the LEDs. The LEDs were
powered by 910µs pulsed currents with two different duty ratios (0.091% and 0.061%) between 0mA and 700mA. The
diode forward voltages corresponding to the short pulsed currents were monitored to correlate junction temperatures
with the forward voltage responses for calibration measurement. In junction temperature measurement, the junction
temperatures at different constant currents and chamber temperature steps were estimated by in-situ monitoring of
forward voltage responses. As a result, we could make realistic estimates of junction temperature to effectively control
conditions for long-term aging tests of high power LEDs.
Index Terms— Light emitting diode, Reliability, Forward voltage, and Junction temperature
1
Introduction
High power LEDs are used for a wide variety of applications, such as display backlighting, communications,
medical services, signage, and general illumination [1][2][3][4][5]. LEDs offer design flexibility; easy adjustment of
line-scale, area-scale, and color dimming; high energy efficiency resulting in low power consumption; possible long
lifetimes up to 50,000 hrs; micro-second-level on-off switching enabling ultrahigh speed response time; wide range of
color temperature; and no low-temperature startup problems.
Despite exciting innovations driven by technological advances and ecological/energy-saving concerns, the LED
industry still faces a challenge in attracting widespread adoption. The concerning issue is the lack of information and
confidence regarding reliability. Data sheets of manufacturers provide user information on optical and electrical
characteristics of high power LEDs at room temperature (or at 25˚C junction temperature). In some cases, thermal
resistance of LED assembly on a printed circuit board is also provided. Absolute maximum rating of junction
temperature on data sheets is typically in the 125˚C to 135˚C range. Users are required to design LED based lighting
systems to maintain the LEDs within this junction temperature guideline.
Use of LEDs at high junction temperature affects internal efficiency, maximum light output, reliability, peak
wavelength and spectral wavelength [6]. For example, a temperature increase is related to forward voltage decrease in
the I-V curve due to a decrease in the bandgap energy of the active region of LEDs as well as a decrease in series
resistance occurring at high temperatures. The resistance decrease is due to higher acceptor activation occurring at
elevated temperatures as well as the resulting higher conductivity of the p-type layer and active layers. When a drive
current is applied to an LED, the LED junction produces heat that needs to be dissipated from the junction to keep the
temperature within limit. A reliable method of estimating junction temperature is necessary to meet the absolute
maximum junction temperature ratings.
Various indirect techniques and models have been attempted to estimate the junction temperature of laser diodes,
including micro Raman spectroscopy, threshold voltage, thermal resistance, photothermal reflectance microscopy,
2
electroluminescence, and photoluminescence [6]. In earlier research conducted, the diode forward voltage has been
used to assess the junction temperature of the pn-junction diode of LEDs [7][8][9]. The forward-voltage method has
two steps, calibration measurement, followed by actual junction temperature measurement. Calibration measurement
by using a pulse current serves as a reference for junction temperature equal to ambient (chamber) temperature and
establishes the relation between the forward voltage and the junction temperature. Actual junction temperature
measurement uses constant current, adding the Joule heating effect to LEDs and derives the relation between the
forward voltage and the junction temperature. In the calibration measurement, a pulsed forward current drives LEDs
[6][7][8] to ensure that the junction temperature is equal to the ambient temperature.
In this paper, we will show how the measurement of junction temperature by forward voltage method can be
performed and improved using tests and measurement instruments, and how this information can then be used to
monitor the operating conditions of high power LEDs. This work will also show that 200 steps of pulse with 0.061%
duty ratio can be used to estimate junction temperature for high power LEDs.
2
Experiment: Devices and Process
For the test using the forward voltage method, 3W high power InGaN LEDs with maximum absolute junction
temperature ratings of 135˚C were mounted on an aluminum metal core printed circuit board (MCPCB). An MCPCB
(LED board) consists of a base layer (aluminum), a dielectric layer (FR-4 layer), and a circuit layer (Cu trace layer) for
higher heat dissipation than the FR-4 board. In each calibration measurement, two LEDs were lighted by a
sourcemeter. Thermocouple wires were attached to the anode side of the lead for the tested LEDs. An overview of the
test setup is shown in Figure 1. In the tests, the LED board was placed in a temperature chamber (Russells chamber RB-16-3-3-LN2) and connected to a sourcemeter (Keithley 2400) generating a short pulse (910µs), a DAQ board (NI
USB 9215) collecting forward voltage responses, and a data logger (Agilent 34970A) monitoring the lead
temperatures. A Labview-controlled computer set was used to program a 910µs short pulse and monitor LED
responses from a Keithley 2400 sourcemeter, NI USB 9215 DAQ board, and Agilent 34970A data logger. Four LEDs
(LED1, LED7, LED9, and LED15) were selected and tested to reflect the location difference in the MCPCB.
Figure 1. Schematic of the Experimental Setup
In the calibration measurement, the LED board was placed into the temperature chamber. The chamber was set at
different temperature steps (20˚C, 35˚C, 50˚C, 65˚C, 80˚C, 95˚C, 110˚C and 125˚C). LED lead temperatures and
chamber temperatures were monitored and recorded every second. 910µs pulses were applied on the LEDs twice in 3
seconds by using a Labview-controlled Keithley 2400 sourcemeter after confirming that the lead temperature of the
LEDs was the same as the chamber temperature (i.e., thermal equilibrium condition). The duty cycle was 0.061% (for
200 peak current levels from 0mA to 700mA) or 0.091% (for a single current level). Voltage responses were stored via
a NI 9215 DAQ board with 300k data in 3 seconds. Chamber temperature was then increased to the next step. The
steps were repeated until the chamber temperature reached 125˚C. In the calibration measurement, a Vf (forward
voltage) vs. Ta (ambient temperature) plot was obtained at the thermal equilibrium condition between the LED
junction and the ambient temperature while avoiding the impact of Joule heating into LEDs using 910µs current pulse
generation.
3
In the junction temperature measurement, junction temperatures at each constant current and chamber
temperature were estimated by the in-situ monitoring of forward voltage responses and lead temperatures of LEDs
every second with a constant test current of 300mA. Chamber temperatures were selected at 20˚C, 35˚C, 50˚C, and
65˚C for 300mA. At each stage of the chamber temperature condition, 30 minutes were needed for lighting the LEDs
to achieve a stable forward voltage. The junction temperature was then estimated by using the previously obtained
ambient temperature (Ta) versus forward voltage (Vf) plot.
3
Experiment Results and Discussion
To reach thermal equilibrium condition between LED junction temperature and ambient temperature, LEDs need
to be placed in a controlled temperature environment into temperature for some moments (i.e., dwell time). The dwell
time of each temperature is around 20 minutes. The dwell time was determined experimentally, and it turned out that
20 minutes was sufficient to reach the thermal equilibrium between the LED package and ambient temperature. Pulse
current was applied when the lead temperature of the LED was equal to the chamber temperature (at thermal
equilibrium) conditions. After the junction has come to thermal equilibrium, a pulsed current of 910µs duration with a
duty ratio 0.091% was sourced into LEDs and the voltage response was measured, as shown in Figure 2. Forward
voltage response from 910µs pulse duration. Three hundred thousand data points were collected in the 3 seconds
needed to detect the output signal and average the voltage responses during the pulse duration.
3.5
Pulse 910µs (duty 0.091%) at 20.3°C & 300mA
Forward voltage (V)
3.0
2.5
2.0
1.5
1.0
0.5
0.0
0.0
0.5
1.0
1.5
2.0
Time (s)
2.5
3.0
Figure 2. Forward voltage response from 910µs pulse duration
To verify pulse duration, a constant current of 700mA (absolute maximum rating from the manufacturer’s data
sheet) was applied until it reached the maximum temperature of lead at an ambient temperature of 21.5˚C. The
ambient temperature was chosen because the change of the lead temperature at 21.5˚C can be shown more clearly than
changes at higher ambient temperatures. The temperature increased linearly until the time reached 0.28 minutes, as
shown in Figure 3. During the initial 0.28 minutes, the lead temperature increased up to 33.8˚C. Therefore,
temperature elevation during 0.28 minutes was calculated as 11.5˚C. The temperature change of the lead at 910µs was
evaluated as 0.00062˚C. If we assume that the junction temperature increases corresponding to the increase of the lead
temperature, a 910µs pulse duration was fairly short to avoid Joule heating from the p-n junction diode of the 3W high
power LEDs.
Further experiments were conducted using this pulse rate of 910µs. Two hundred pulses were applied at each
chamber temperature with a 0.061% duty cycle by using a single pulse in each step, as shown in Figure 4. I-V curves
obtained from 200 steps of single pulse generated by 910us pulse and 0.067% duty cycle. Without storing the Joule
heating effect of LEDs, the forward voltages at different ambient temperatures as reference values can be used for
estimating the junction temperature of LEDs at any specific current among 200 different current levels from 0mA to
700mA. Figure 5 shows Vf (forward voltage) vs. Ta (ambient temperature) plots obtained linearly from these
multi-steps of 910µs pulse and chamber temperature (20˚C, 35˚C, 50˚C, 65˚C, 80˚C, 95˚C, 110˚C, and 125˚C). Two
hundred steps of the I-V pulse curve also showed that when the ambient temperature (the same as the junction
temperature) was higher, the forward voltage response went lower and the slope in I-V curve became steeper.
4
42.5
40.0
Temperature (°C)
37.5
Lead temperature of LED9
Linearly increased section (0.28 minutes)
35.0
32.5
30.0
27.5
25.0
22.5
20.0
0
3
6
9
12 15 18 21 24 27 30
Time (min.)
Figure 3. Pulse duration verification at 25˚C ambient temperature with 700mA constant current
LED1 20°C
LED1 35.1°C
LED1 49.8°C
LED1 65.4°C
LED1 80.4°C
LED1 95°C
LED1 109.9°C
LED1 125.3°C
0.8
0.7
Current (A)
0.6
0.5
0.4
0.3
Increasing temperature
0.2
0.1
0.0
2.4
2.6
2.8
3.0 3.2 3.4
Voltage (V)
3.6
3.8
4.8
4.6
4.4
4.2
4.0
3.8
3.6
3.4
3.2
3.0
2.8
2.6
100mA to LED7
200mA to LED7
300mA to LED7
400mA to LED7
500mA to LED7
600mA to LED7
700mA to LED7
20 30 40 50 60 70 80 90 100 110 120 130
Chamber temperature (°C)
Forward voltage (V)
Forward voltage (V)
Figure 4. I-V curves obtained from 200 steps of single pulse generated by 910us pulse and 0.067% duty cycle
100mA to LED15
200mA to LED15
300mA to LED15
400mA to LED15
500mA to LED15
600mA to LED15
700mA to LED15
4.8
4.6
4.4
4.2
4.0
3.8
3.6
3.4
3.2
3.0
2.8
2.6
20 30 40 50 60 70 80 90 100 110 120 130
Chamber temperature (°C)
Figure 5. Forward voltage vs. chamber temperature of LED7 and LED15
The values of slopes and intercepts in the duty ratio 0.091% were compared with those in the duty ratio of 0.061%,
as shown in Table 1. It turned out that the average values and standard deviations of the slope a2 and intercept b2 from
300mA result obtained by 200 steps of 910µs pulse with 0.061% duty are similar to average values and standard
deviations of the slope a1 and intercept b1 from a single current level (at 300mA) with 0.091% duty ratio. This result
confirmed that 200 steps of pulse with a 0.061% duty ratio can be practically used to construct a reference plot of Tj vs.
Vf at a specific current level between 0mA to 700mA. The minor difference was caused by the variation of chamber
temperature (0.4˚C) at each step of ambient temperature during the experiments.
5
Table 1. Comparison of slope and intercept at 300mA between 0.091% duty cycle test and 0.061% duty cycle test result
Tj = a1 + b1 × Vf
Duty cycle 0.091% at 300mA
a1
LED1
1655.9
LED7
1881.4
LED9
1535.6
LED15
1509.1
a1
Average
1645.5
Standard deviation
169.8
Tj = a2 + b2 × Vf
Duty cycle 0.061% at 300mA
a2
LED1
1661.8
LED7
1869.0
LED9
1539.3
LED15
1515.1
a2
Average
1646.3
Standard deviation
161.8
b1
-506.0
-580.9
-463.6
-458.6
b1
-502.3
56.6
b2
-507.1
-576.1
-463.9
-459.8
b2
-501.7
54.0
In Figure 6, the voltage responses at constant current 300mA show that the Joule heating effect changes the
voltage characteristics of LEDs. A constant current of 300mA is useful to combine the high ambient temperature
effect and the Joule heating effect from LEDs. The test results of junction temperature measurement at a constant
current of 300mA demonstrated that the junction temperature reached the maximum absolute rating of 135˚C at
around 65˚C chamber temperature and 80˚C of lead temperature, as shown in Figure 7.
LED1 at constant current 300mA
LED7 at constant current 300mA
LED9 at constant current 300mA
LED15 at constant current 300mA
3.11
3.10
3.09
Forward voltage (V)
3.08
3.07
3.06
3.05
3.04
3.03
3.02
3.01
3.00
15
20
25
30 35 40 45 50 55
Chamber temperature (°C)
60
65
70
Lead temperature (°C) Junction temperature (°C)
Figure 6. Forward voltage response at constant current 300mA
LED1 junction at 300mA
LED7 junction at 300mA
LED9 junction at 300mA
LED15 junction at 300mA
150
140
130
120
110
100
90
80
20
25
30
35
40
45
50
55
60
65
55
60
65
Chamber temperature (°C)
80
70
60
50
40
20
25
30
35
40
45
50
Chamber temperature (°C)
Figure 7. Junction temperature measurement at constant current 300mA
6
4
Conclusions
In this study, junction temperatures of 3W high-power LEDs mounted on an aluminum metal core printed circuit
board were estimated at various ambient temperatures and drive currents with 910µs pulsed currents and a duty ratio of
0.061%/ 0.091%. A pulse duration of 910µs was able to generate short-pulse current without generating Joule heating
effects from the LED die. The test results showed that 200 steps of pulse with 0.061% duty ratio can be used
practically to construct a reference plot of junction temperature (Tj) vs. forward voltage (Vf) at a specific current level
between 0mA and 700mA for 3W high power LEDs. From the relationship between junction temperature and the
forward voltage, the junction temperature of high power LEDs was directly estimated to control conditions for
long-term aging tests and the reliability of high power LEDs.
This approach shows a direct way of estimating junction temperature by considering forward voltage and junction
temperature. It doesn’t need to use temperature data of reference points those in the previous thermal resistance
approach. Also, once results of multi-current steps for all range of operating currents are obtained, junction
temperature at specific current and ambient temperature can be derived without experiments at each single current.
The previous approach could not obtain junction temperatures at different current levels as it derived junction
temperature at a single current. This result helps LED designers and customers to estimate junction temperature
without the uncertainties that exist in current methods.
5
Acknowledgments
The authors would like to thank the more than 100 companies and organizations that support research activities at
the Center for Advanced Life Cycle Engineering at the University of Maryland annually, specifically the CALCE
Prognostics and Health Management Group. Also, the authors would like to thank Shivangi Parikh and Mark
Zimmerman for copyediting.
6
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
M. R. Krames et al., “Status and Future of High-Power Light-Emitting Diodes for Solid-State Lighting”,
Journal of Display Technology, vol. 3, no. 2, pp. 160-175, 2007.
D. A. Steigerwald et al., “Illumination with Solid State Lighting Technology”, IEEE Journal on
Selected Topics in Quantum Electronics, vol. 2, pp. 310-320, 2002.
F. M. Steranka et al., “High Power LEDs - Technology Status and Market Applications”, Physica Status
Solidi (a), vol. 194, pp. 380-388, 2002.
Y. Aoyama and T. Yachi, “An LED Module Array System Designed for Streetlight Use”, Energy 2030
Conference, 2008. Energy 2008. IEEE, pp. 1-5, 2008.
R. Vittori and A. Scaburri, “New Solid State Technologies and Light Emission diodes as a Mean of
Control and Lighting Source Applicable to Explosion Proof Equipment, With the Scope to Reduce
Maintenance, to Limit the Risk of Bad Maintenance and to Expand the Plants’ Life”, PCIC Europe,
2009. PCIC EUROPE ΄09. Conference Record, pp. 193-198, 2009.
S. Chhajed et al., “Junction Temperature in Light-Emitting Diodes Assessed by Different Methods”,
Progress in Biomedical Optics and Imaging – Proceedings of SPIE, vol. 5739, Light-Emitting Diodes:
Research, Manufacturing, and Applications IX, pp. 16-24, 2005.
J. Chonko, “Using Forward Voltage to Measure Semiconductor Junction Temperature”, Keithley
Instruments, Inc., no. 2681, pp. 1-3, 2006.
Y. Xi and E. F. Schubert, “Junction-temperature Measurement in GaN Ultraviolet Light-emitting
Diodes Using Diode Forward Voltage Method”, Applied Physics Letters, vol. 85, no. 12, pp. 2163-2165,
2004.
F. Jiang et al., “Research on the Junction-temperature Characteristic of GaN Light-emitting Diodes on
Si Substrate”, Journal of Luminescence, vol. 122-123, pp. 693-695, 2007.