In the framework of EUROMET project No. 353 "Transmittance of V(l) intercomparison", a filter with a transmittance close to the V(l) function of the CIE photometric standard observer has been circulated between five participating laboratories. The spectral transmittance of the central part of the filter was measured at filter temperature 25 ±1 °C. The objective of the intercomparison was to confirm the estimated uncertainty (filter part) for each participating laboratory when using a calibrated V(l) filter together with a precision aperture and a silicon trap detector for realization of photometric scales.SP measured the transmittance of the filter before and after circulation to each participant.
The V(l) filter chosen in this comparison is intended to be used for accurate illuminance measurement together with a silicon trap detector. Therefore the most important quantity for this intercomparison is the integrated value E, of the spectral transmittance t(l) over the visible spectra range weighted with the spectral responsivity of the trap detector, i.e. a linear dependency with the wavelength l, and with the relative spectral irradiance, Sl(l) of a luminous intensity source which is close to the CIE standard source A, 2856 K. where a is a constant. In the Table below this equation is applied on t(l) measured at the different institutes and the results are divided by the corresponding SP value. For two labs also the effect of wavelength corrections above are applied.
Table. Spectrally weighted transmittances relative SP (ELab/ESP)
|Lab 1||Lab 2||Lab 3||Lab 4|
The result in the Table above shows good agreement, within 0,1 %, between four of the participating laboratories. The disagreement with the fifth institute was quite large. To some extent this could be explained by the fact that the transmittance of the filter decreased, about 1 %, after the year 2002.Wavelengths corrections, visually improving the spectral differences in transmittance for two of the participators, improved also the spectrally weighted transmittance for one of them but not for the other. The reason for this is that wavelengths near the maximum of the V(l)-function are more important than others.With exception for a 1 % transmittance drop 2002, the filter transmittance showed a very good stability during the comparison period 1995-2003. During this period the filter was not cleaned.Full report can be asked for from Mr Leif Liedquist. Intercomparison of spectral transmitance on temperature stabilized V(lambda) filters which are used in filter radiometers for realization of a photometric scale. Also neutral density filters are included in the intercomparison.