Search

A new interferometer for calibration of graduated line scales is described. It uses a novel method for observation of the distance between the graduation lines. The line images are recorded by a moving microscope and a CCD camera. A frame-grabber unit is used to digitize the images. We calculated the position of the microscope at the moment of charging the video fields by the use of digitized interference and video-synchronization signals. For elimination of the effect of carriage rotations, the focus point of the microscope and the apex of the main-arm cube-corner reflector are adjusted to the same point. The measurement and the analysis processes are completely automated. The estimated overall uncertainty (1[Sigma]) for a 1-m Invar scale is 44 nm.

The construction of a passively stabilized external cavity diode laser operating at 780 nm is reported. The sensitivity of laser frequency to changes in air pressure was studied and subsequently eliminated. The relative frequency stability obtained was 4 x [10.sup.-9] for an integration time of 4000 s.

Array spectroradiometers are attractive alternatives to scanning spectroradiometers in solar ultraviolet measurements. However, the measurement of solar spectral irradiance imposes stringent requirements for the linearity of the instruments. In this article, two array spectroradiometers were characterized for nonlinearity. Significant nonlinearities, in excess of 10%, as a function of analog-to-digital converter counts were discovered. Additional nonlinearities as a function of integration time were observed at very long integration times. No clear residual nonlinearity as a function of spectral irradiance was witnessed despite the characterization spanning four orders of magnitude of spectral irradiance. The characterizations were carried out with three measurement setups that are briefly compared.

Microbridges are miniature suspended structures fabricated in silicon. Passing a current through the microbridge can heat it up to the point of incandescence. A glowing microbridge can be used as a wideband light source. This study presents a method for optical measurement of the temperature of a microbridge. Spectroscopic measurements of microbridges are optically challenging, because the multilayer structures cause interference effects. To determine the temperature from the emitted spectrum, the emissivity was modeled with thin-film Fresnel equations. Temperatures of 500-1100 degrees C were obtained from the measured spectra at different levels of applied power. The range is limited by the sensitivity of the detectors at lower power levels and by the stability of the bridge at higher levels. Results of the optical measurements were compared with contact temperature measurements made with a microthermocouple in the same temperature range. The results of the two methods agree within 100 K.

We have developed a physical model for the spectral irradiance of 1 kW tungsten halogen incandescent lamps for the wavelength range 340-850 nm. The model consists of the Planck's radiation law, published values for the emissivity of tungsten, and a residual spectral correction function taking into account unknown factors of the lamp. The correction function was determined by measuring the spectra of a 1000 W, quartz-halogen, tungsten coiled filament (FEL) lamp at different temperatures. The new model was tested with lamps of types FEL and 1000 W, 120 V quartz halogen (DXW). Comparisons with measurements of two national standards laboratories indicate that the model can account for the spectral irradiance values of lamps with an agreement better than 1% throughout the spectral region studied. We further demonstrate that the spectral irradiance of a lamp can be predicted with an expanded uncertainty of 2.6% if the color temperature and illuminance values for the lamp are known with expanded uncertainties of 20 K and 2%, respectively. In addition, it is suggested that the spectral irradiance may be derived from resistance measurements of the filament with lamp on and off.

A comparison between the absolute gonioreflectometric scales at the Helsinki University of Technology (TKK) and the Physikalisch-Technische Bundesanstalt (PTB) has been accomplished. Six different reflection standards were measured for their 0:45 spectral radiance factor between 250 and 1650 nm in 10 nm intervals. Also, the 0:d reflectance factor between 400 and 1600 nm in 100 nm intervals was determined from goniometric reflectance measurements over polar angles with subsequent integration within the hemisphere above the sample. Goniometric comparisons covering such an extensive wavelength range and also several different sample materials are rarely implemented. For all but one sample, the difference between the results obtained at the TKK and the PTB was, with the exception of a couple of measurement points, within the expanded uncertainty (k = 2) of the comparison at least up to a wavelength of 1400 nm. All differences between the measurement results can be understood, except for one translucent sample in the visible wavelength range. The effect of sample translucency was found to be significant in the NIR wavelength region. Also, a general tendency of an increase of the TKK values relative to the PTB values in the UV region was observed. Possible causes for this phenomenon are discussed.

The energy transfer integral between radiating rectangular and detecting circular parallel plates having nonideal angular characteristics is solved for modeling the distance dependence of the irradiance signal. The equation derived for the irradiance signal, which is called the modified inverse-square law, depends on the position, shape, size, and angular characteristics of the light source and the detector. We apply the new model equation to the calibration of a spectroradiometer to determine accurately the distance offsets, which fix the positions of the effective receiving apertures of diffusers used in the entrance optics of spectroradiometers. Earlier measurement results, e.g., for solar UV irradiance, may include uncorrected effects and can be corrected reliably as diffuser offsets and other correction factors are determined with the modified inverse-square law. Simplifications of the modified inverse-square law for analyzing the distance offsets and the correction factors are studied. Simplified equations for the diffuser offset analysis may be used without losing the accuracy when the cosine response of the diffuser is reasonably good. However, for diffusers whose angular responsivities deviate much from the cosinusoidal angular responsivity, large approximation errors in the diffuser offset values may appear if the angular effects are not properly taken into account.

A goniofluorometer has been built that is capable of measuring in various viewing angles ranging from 10 degrees to 90 degrees . The incident angle can be varied from 0 degrees to 8 degrees . The goniofluorometer can measure bispectral luminescent radiance factors in the wavelength range of 250-800 nm. To our knowledge, there are no other reported results of similar devices capable of spectral measurements in various measurement geometries.

Independent methods for measuring the absolute spectral irradiance responsivity of detectors have been compared between the calibration facilities at two national metrology institutes, the Helsinki University of Technology (TKK), Finland, and the National Institute of Standards and Technology (NIST). The emphasis is on the comparison of two different techniques for generating a uniform irradiance at a reference plane using wavelength-tunable lasers. At TKK's Laser Scanning Facility (LSF) the irradiance is generated by raster scanning a single collimated laser beam, while at the NIST facility for Spectral Irradiance and Radiance Responsivity Calibrations with Uniform Sources (SIRCUS), lasers are introduced into integrating spheres to generate a uniform irradiance at a reference plane. The laser-based irradiance responsivity results are compared to a traditional lamp-monochromator-based irradiance responsivity calibration obtained at the NIST Spectral Comparator Facility (SCF). A narrowband filter radiometer with a 24 nm bandwidth and an effective band-center wavelength of 801 nm was used as the artifact. The results of the comparison between the different facilities, reported for the first time in the near-infrared wavelength range, demonstrate agreement at the uncertainty level of less than 0.1%. This result has significant implications in radiation thermometry and in photometry as well as in radiometry.

The optical parameters of a SiO2 thin-film coating determined from the spectral reflectance and transmittance measurements at various incidence angles, including the normal incidence and the Brewster's angle, are compared in this paper. The high-accuracy measurements were carried out through visible-near-infrared spectral regions by using our purpose-built instruments. The optical parameters obtained from the reflectance and the transmittance data are consistent over the angles of incidence and agree within 0.2%. The effect of important systematic factors in the oblique-incidence spectrophotometric measurements is also discussed.

It is difficult to predict where the effective measurement plane is situated with dome-shaped diffusers often used in commercial photometers and radiometers. Insufficient knowledge of this plane could lead to large systematic errors in calibration of the illuminance responsivity of photometers. We propose a method that can be used to determine this reference plane accurately, based on the inverse-square law between the measured signal and the distance from the source. The method is demonstrated with three commercial photometers with dome-shaped diffusers of different geometries. By taking into account the measured shifts of the reference planes (5.0 +/- 0.5 mm, 7.8 +/- 0.3 mm, and 8.5 +/- 0.7 mm), we reduced the systematic measurement errors up to 2% to statistical uncertainty components at the level of 0.2%.

Gonioreflectometric determination of reflectance factors that involves hemispherical collection of reflected flux, which is an alternative to integrating sphere-based methods, is discussed. A detailed description of a gonioreflectometer built at the Helsinki University of Technology is presented. The instrument is used to establish an absolute scale of total diffuse reflectance factors throughout the spectral range 360-830 nm. The hemispherical reflectance factors are obtained through integration of the gonioreflectometric measurement results. The reflectance factors of white high-quality artifacts can be determined with a combined standard uncertainty of 0.20%. Results of test measurements were found to be in agreement with values traceable to other absolute scales based on integrating-sphere methods.

We improve the methods used to interpolate the responsivity of unbiased silicon photodetectors in the near-ultraviolet region. This improvement is achieved by the derivation of an interpolation function for the quantum yield of silicon and by consideration of this function in the interpolation of the internal quantum efficiency of photodiodes. The calculated quantum-yield and spectral-responsivity values are compared with measurement results obtained by the study of a silicon trap detector and with values reported by other research groups. The comparisons show agreement with a standard deviation of 0.4% between our measured and modeled values for both the quantum yield and the spectral responsivity within the wavelength region from 260 to 400 nm. The proposed methods thus extend the predictability of the spectral responsivity of silicon photodetectors to the wavelength region from 260 to 950 nm. Furthermore, an explanation is proposed for the change in the spectral responsivity of silicon photodiodes that is due to UV radiation. In our improved quantum efficiency model the spectral change can be accounted for completely by the adjustment of just one parameter, i.e., the collection efficiency near the SiO(2)/Si interface.

We have developed a new method for characterizing the irradiance responsivity of filter radiometers. The method is based on a spatially uniform, known irradiance, generated by combining several identical laser beams. The measurement setup and the experimental demonstration at one wavelength are presented. The diffraction correction related to the generated irradiance is studied experimentally. The uncertainty analysis of the method indicates a relative standard uncertainty of 1 x 10(-3). The results with the new method are compared with the characterization measurements based on our present spectral-irradiance scale. The results have a relative deviation of 1 x 10(-3), which is well within the combined standard uncertainty of the comparison.

Nonlinearities of the responsivity of various types of siliconphotodetectors have been studied. These detectors are based onphotodiodes with two sizes of the active area (10 x 10 mm(2) and 18 x 18 mm(2)). The detectorconfigurations investigated include single photodiodes, two reflectiontrap detectors, and a transmission trap detector. For all devices, the measured nonlinearity was less than 2 x 10(-4) forphotocurrents up to 200 muA. The diameter of themeasurement beam was found to have an effect on thenonlinearity. The measured nonlinearity of the trap detectorsdepends on the polarization state of the incident beam. Theresponsivity of the photodetectors consisting of the large-areaphotodiodes reached saturation at higher photocurrent values comparedwith the devices based on the photodiodes with smaller activearea.

A precision spectrometer was used to measure the spectral reflectance of a silicon photodiode over the wavelength range from 250 to 850 nm. The results were compared with the corresponding values predicted by a model based on thin-film Fresnel formulas and the known refractive indices of silicon and silicon dioxide. The good agreement at the level of 2 x 10(-3) in the visible wavelength range verifies that the reflection model can be used for accurate extrapolation of the spectral reflectance and responsivity of silicon photodiode devices. In addition, characterization of the photodiode reflectance in the ultraviolet region improves the accuracy of the spectral irradiance measurements when filter radiometers based on trap detectors are used.

A detector-based absolute scale for spectral irradiance in the 380-900-nm wavelength region has been developed and tested at the Helsinki University of Technology (HUT). Derivation of the scale and its use for photometric and colorimetric measurements are described. A thorough characterization of a filter radiometer, constructed from a reflection trap detector, a precision aperture, and a set of seven temperature-controlled bandpass filters, is presented. A detailed uncertainty analysis of the scale indicates a relative standard uncertainty of approximately 0.2% throughout most of the wavelength region. The standard uncertainties obtained in measurements of correlated color temperature and luminous intensity of three Osram Wi41/G tungsten-halogen lamps are 2 K and 0.3%, respectively. The spectral irradiance scale is compared with the HUT luminous intensity scale. The agreement of the results at the 0.1% level is well within the combined standard uncertainty of the two scales.

A six-element polarization-independent transmission trap detector with coaxial input and output beams has been constructed and full characterized. The measured optical parameters are compared with their values, predicted by Fresnel equations. Measured transmittances are in agreement with the predicted values within 2 x 10(-5) in the wavelength region from 450 to 650nm. The spectral responsivity of the transmission trap detector is in agreement with the predicted values within 0.035% at 543.5- and 633.0-nm vacuum wavelengths. The spatial uniformity of the responsivity is +/-0.03% across the active area of approximately 5 x 6 mm(2), measured with a laser beam of 1-mm diameter. The angular uniformity of the transmission trap detector is better than +/-0.01% for +/-3 degrees rotation around two perpendicular axes.

A single-mode GaAlAs laser is frequency stabilized by keeping the injection current and the voltage across the laser constant. A long-term frequency stability of 7 x 10(-8) and a reproducibility of 3 x 10(-7) were obtained.