Search

Recent studies in the epidermis have shown that Far-UVC (200-230nm) is a promising candidate against Novel Coronavirus (SARS-Cov-2) with little DNA damage. Due to the consideration that conventional Far-UVC KrCl excilamps may emit 200-230 nm radiation (typically 222-nm peak wavelength) but with some harmful UV radiation beyond 230 to 280 nm, a novel design of Far-UVC KrCl excilamps with the filter and reflector is introduced to reduce the harmful UV radiation from 10.9% to 2.5% at the cost of 30%~40% reduction in the total irradiance. In our study, the radiant characteristics and service life of the novel Far-UVC KrCl excilamps of 40~75 Watt (electrical power) with 222-nm peak wavelength were investigated. The service life was assessed under aging at the ambient temperatures (Ta) of 25 and 85 for 500 hours, respectively. The results showed that both the ambient temperature and the root mean square of current (Irms) into the excilamps have a substantial effect on the lifetime of the KrCl excilamps. Furthermore, although no significant change of the off-nominal emission ratio existed during the lifetime test, it was observed that the high ambient temperature has a negative effect on the filtering of the harmful radiation. © 2023 IEEE.

The lifetime projection of light-emitting diodes (LEDs) driven by DC current has been addressed by IESNA LM80-80 and TM-21-11. For applications such as visible light communication (VLC) and backlight of displays, the lifetime estimation of LEDs strobed with the pulsed current at various frequencies and duty cycles is in demand. Thus, the lifetime and failure analysis of LEDs driven by constant-amplitude current with pulse width modulation (PWM) is investigated in the present study. It is found that the pulsed loading may introduce more localized current crowding with overheating than the DC current. The induced thermal stress may result in luminous decay and catastrophic failures such as the damage in the current spreading layer and the bonding wire. The lifetime with respect to the peak junction temperature, defined by the highest junction temperature when LEDs driven by the alternating PWM current reach their steady state, is analyzed and presented. It is concluded that the lifetime of LEDs may be divided into 3 regions: the Luminous Decay (LD) region, the Catastrophic Failure (CF) region and the LD+CF region. In the LD region, the projection of lifetime may follow TM-21-11. In the CF region, the sudden failure without significant luminous decay may occur. In the LD+CF region, although lifetime can be projected by LD, the CF may appear before the lifetime projection. © 2022 IEEE.

Optical characteristics and service life of the Germicidal-UVC (GUVC) low-pressure Hg lamp and the Far-UVC KrCl excimer lamp (FUVC excilamp) were investigated. The results showed that UV irradiance at the center of the FUVC excilamp and the GUVC lamp was 0.22 mW/cm2 and 9.27 mW/cm2 at a working distance of 5 cm, respectively. To achieve the same virus inactivation effect, the harmful power in range of 230280 nm should be more than 20 times that of the FUVC excilamp without considering the UVA (320400 nm) and UVB (280320 nm). Furthermore, the service life was assessed under room temperature 30 ± 5 °C. It was found that the average service life at L50 (with the irradiance decay of 50%) of the FUVC excilamp and GUVC lamp was about 2, 600 and 4, 700 hours, respectively. © 2021 IEEE.

The KrCl Excimer lamps (Excilamps) with 222nm peak wavelength have been developed for 20 years, however, there are few researches found regarding on their life projection. To insure the enough radiation for the important inactivating effectiveness during the service life, this work focuses on irradiance maintenance of KrCl Excilamps at different tube wall temperatures. In general, the activation energy of the Arrhenius equation modelling the lifetime for electronic products suffering from temperature stress below 500K can be considered as temperature-independent. However, due to the unique configuration of KrCl Excilamps without the electrodes inside their tubes and the chemical reaction of Chlorine with the quartz tube wall, their degradation model might need to be verified. In our study, the irradiance of the KrCl Excilamps at tube temperatures around 80, 90 and 100°C for 1000 hours was investigated to find the associated rate coefficients using the exponential and the logarithmic least square curve-fitting. For the better accuracy of life projection, the power-law and the linear relations with temperature are assumed to obtain the temperature-dependent Arrhenius parameters: the pre-exponential factor and the activation energy of the Arrhenius Equation. It is found that the life projection based on the irradiance maintenance can be obtained with a better agreement by the temperature-dependent models using the logarithmic least square curve-fitting for the KrCl Excilamps with 222nm peak wavelength. © 2021 IEEE.

The measurement of the junction temperature of light emitting diodes (LEDs) has been crucial for the thermal management. It is more important for the applications of the ultra-violet light emitting diode (UVLED) with a low efficiency generating more heat and with the PWM-mode current driving at a higher peak current than the one with the DC-mode current driving. In this paper, the dynamic junction temperature of the UVLED with 365nm peak wavelength is characterized by the empirical modelling for the PWM-mode current driving modes with different duty cycles and periods using T3ster with the time-resolved measuring based on the JESD 51-14 and 51-51 for the DC-mode current driving. The PWM-mode response of the junction temperature rise and its fluctuation are normalized by the DC-mode response and its PWM-mode response respectively in terms of duty cycles (ranged from 10~90%) and periods (ranged from 0.1msec. to 1sec.). The DC-mode junction temperature rise with respect to applied dc currents can be direct measured by T3ster. Based on the DC correlation, the PWM-mode correlations with the power-exponential function of the normalized peak junction temperature rise and its fluctuation is established. With the proposed correlations, the normalized mean junction temperature rise as the average of the dynamic junction temperature rise over a period is found. The normalized thermal resistances of the peak, mean, fluctuation are inferred from the correlations. Furthermore, the maximum permissible pulse current is able to be determined based on the maximum rating of the DC junction temperature and the consideration of current crowding. Finally, the pulse density modulation (PDM) mode of current driving is also discussed. Those successful empirical modeling is helpful for the thermal design and thermal management for applications of UVLED with PWM-mode driving. © 2020 IEEE.

Infrared (IR) and thermocouple (TC) measurements of the light emission surface of white, blue, red and green LEDs are demonstrated using Transient Thermal Tester (T3Ster). It is shown that the IR measurement could obtain a more accurate result which is lower than the junction temperature (T-j) of LED. However, the TC measurement led to a higher temperature than T-j due to the heating effect resulted from the radiation of LED. IR measurement of the light emitting surface temperature of LEDs with dispensing and remote phosphor driven by the pulsed current is analyzed to show the transient temperature response on light emission surface with correlation between the peak light emitting surface temperature and the pulse current and width.

An IR vertical-cavity surface-emitting laser (VCSEL) with a peak wavelength 940nm for ToF/Flood applications is characterized by the empirical modelling. This paper presents a novel approach to predict pulsed junction temperature rise of VCSELs using T3ster to extend the time-resolved measuring with a constant bias current based on the JESD51-14. For different DC bias currents, it is found the normalized junction temperature rise and drop curves with a reflectional symmetry suggest the same thermal model for simplifying the measurement. For the pulse bias, the thermal behavior of VCSEL due to frequency, duty and pulse number is studied. And an algorithm predicting the pulsed junction temperature rise and fluctuation were proposed and showing a good consistent result with the measurement of the T3ster at the first few pulses with low duties. The result also shown the pulsed optical power output was reduced due to the increase of the junction temperature rise in the pulse duty.

The junction temperature of LEDs is important for its life, reliability and efficacy. The existing transient measurement method of the junction temperature of using transient thermal tester (T3ster) based on the time-resolved measuring with a constant bias current according to the JESD51-14 is able to obtain the LED junction temperature rise based on the known temperature of thermostat module as the heatsink. In practice, without T3ster and the thermostat module, for multi-LED module with multiple heat sources in the luminaries, the thermal resistance matrix needs to be built to realize the thermal transfer between one another such that their junctions temperature raises can be calculated from their thermal power consumption. It is demonstrated that without measurement of the temperature heatsink temperature, the individual junction temperature of LEDs (groups) in multi-LED module with a thermistor (NTC or PTC) on its PCB measuring the local reference temperature can be obtained from the modified N x (N+1) thermal resistance matrix having N LED heat sources and N+1 junction temperature rise. A multi-LED module sample with 3 groups of LEDs has been studied at 5 different ambient temperatures and 2 additional different power combinations using 3 thermal resistance matrices built at 3 different ambient temperatures for comparison. The result showed that the modified method is effective and convenient for measurement of the junction temperature of the multi-LED modules with a similar accuracy comparing to the conventional one in the local reference temperature range of 20~100°C.

This paper presents dynamic thermal analyses of LEDs for the flash unit of a camera driven by current pulses. A new approach with the modified T3ster has been successfully developed to extend the time-resolved measuring with a constant driving current based on the JESD51 to dynamic junction temperature with pulsed current. For comparison, CAD simulation using Flotherm was performed. Good agreement between experimental measurement and computational modelling was achieved.

In recent years, LEDs are widely adopted as a new light source in various applications. Among the many advantages of LEDs over traditional light sources, one key advantage is its long lifespan which substantially reduces the overall maintenance cost. Nevertheless, such characteristic causes various difficulties in the prediction of lifespan and reliability assessment. The lifespan of LEDs can be predicted by a proper reliability test. One common method is to monitor the luminous flux maintenance of LEDs operating under different temperatures. The luminous flux is measured at different time intervals. A prediction model is then used to predict the lifespan of LEDs based on the data collected. In order to obtain an accurate prediction, both the luminous flux and the elapsed time are required to he recorded precisely during the reliability test. Since the luminous flux drop in the first few thousand hours is insignificant, it is common to record only a few percent drop throughout the entire test. This drop level is comparable to the measurement uncertainty. The accuracy of the lifespan prediction will be affected in particular when the error of the measurement system is significant. In this paper, a sensitive study of the data acquisition error on lifespan prediction was carried out. The confidence band was very sensitive to the luminous output measurement error, which usually comes from the periodic optical characterization during the reliability test. This error can be eliminated by in-situ monitoring of the illuminance of the samples in the chamber throughout the reliability test.