At present day, in the field of lighting the incandescent lamps are phasing out. The solid state lighting products, i.e. LED, and the related market are developing very fast in China for its promising application, due to the energy-saving and the colorful features. For the quality control and the commercial trade purpose, it is highly necessary to measure the optical parameters of LED light sources with a fast, easy and affordable facility. Therefore, more test labs use the spherical spectrometer to measure LED. The quasi- monochrome of LED and the V(lambda) of silicon photodetector mismatch problem is reduced or avoided, because the total spectral radiant flux (TSRF) is measured, and all the optical parameters are calculate from the TSRF. In such a way, the spherical spectrometer calibration requires TSRF standard lamps instead of the traditional total flux standard lamps. National Institute of Metrology China (NIM) has studied and developed the facilities for TSRF measurement and provides related calibration services. This paper shows the TSRF standard lamp calibration procedure using a spherical spectrometer in every-day calibration and its traceable link to the primary SI unit at NIM. The sphere is of 1.5 m diameter, and installed with a spectrometer and a silicon photodetector. It also shows the detail of data process, such as the spectral absorption correction method and the calculation of the result derived from the spectral readings. The TSRF calibration covers the spectra range of 350 nm to 1050 nm, with a measurement uncertainty of 3.6% ~ 1.8% (k=2).
Averaged LED Intensity was defined in CIE 127, in which two measurement conditions were proposed, both required the area of entrance aperture to be 100 mm<sup>2</sup>. The CIE 127 LED measurement method is widely accepted. Circular entrance aperture is usually used in commercial photometers, but the photosurface of silicon photodetector is square. However, when LEDs with a narrow beam angle are measured their Averaged LED Intensity, the result may vary due to various aperture shapes. In this paper, a mathematical measurement model is built, then the uncertainty of measurement introduced by shape of aperture is rigorously evaluated by comparing the prediction of model and the results obtained from measurement experiments. It is an effective quantitative analysis to the impact of aperture shapes on accurate measurement of Averaged LED Intensity.
Using a goniophotometer to implant a total luminous flux measurement, an error comes from the sampling interval, especially in the situation for LED measurement. In this work, we use computer calculations to estimate the effect of sampling interval on the measuring the total luminous flux for four typical kinds of LEDs, whose spatial distributions of luminous intensity is similar to those LEDs shown in CIE 127 paper. Four basic kinds of mathematical functions are selected to simulate the distribution curves. Axial symmetric type LED and non-axial symmetric type LED are both take amount of. We consider polar angle sampling interval of 0.5°, 1°, 2°, and 5° respectively in one rotation for axial symmetric type, and consider azimuth angle sampling interval of 18°, 15°, 12°, 10° and 5° respectively for non-axial symmetric type. We noted that the error is strongly related to spatial distribution. However, for common LED light sources the calculation results show that a usage of polar angle sampling interval of 2° and azimuth angle sampling interval of 15° is recommended. The systematic error of sampling interval for a goniophotometer can be controlled at the level of 0.3%. For high precise level, the usage of polar angle sampling interval of 1° and azimuth angle sampling interval of 10° should be used.