Physics of auroral phenomena : proceedings of the 36th Annual seminar, Apatity, 26 February – 01 March, 2013 / [ed. board: A. G. Yahnin, A. A. Mochalov]. - Апатиты : Издательство Кольского научного центра РАН, 2013. - 215 с. : ил., табл.

Practice o f CCD cameras' calibration by LED low-light source The results of the measurements for both cameras are presented in Fig. 2. The lines 1, 2, 3 are the light intensity (in counts per second) registered with ‘Red’, ‘Green’ and ‘Blue’ filters, correspondently, after subtraction o f the black level (line 4) obtained for the same region in the CCD matrix. The error bars illustrate the standard deviation o f the averaged over the 10 frames value. One can see that the measured curves are perfectly linear in this reference frame, which means the parameterization of the lines is: Cobig) - CBf{g) = BjQxpikjg), (3) whereZ^ and k ; are the constants presented in Table 1. For wavelengths o f ‘Blue’ and ‘Green’ filters the ^corresponds well to value -0.035 dB/step given by manufactors, while for ‘Red’ filter the value is -10% less. Table 1.Parameterization o f the lines in Fig. 2 by 3. Tests of the light source SID 105 The light source SID 105 was assembled from LED- based flashlight and matt glasses in a tube. The flashlight has 5 intensity settings: 0 - off, 1 - low, 2 - medium, 3 - max, 4 - blinking. During 3 October, 2012 in Apatity we made a set of measurements with the SID 105 powered from several sources. Starting from usual set o f batteries we also tested an external laboratory supply with stabilization of voltage. There were two aims for the measurements: i) to follow simultaneously DC current-voltage characteristic of the SID 105 lamp and light intensity during -3 0 minutes discharge o f new alkaline batteries (the same batteries that was used for measurements by cameras G-l and G-2 described above); ii) to check proportionality o f the light intensity in the selected spectral regions to the DC current through the light source. Then we hope to deduce the correction factors needed due to difference in the DC currents during different calibration experiments. The measurements with battery power source give us estimations o f the battery parameters during camera calibration described in previous section. Simultaneous measurements o f electric and emission parameters o f the SID 105 forthe intensity settings 1,2 and 3 were performed with stable stationary power source. As a summary of the measurement we can conclude: i) for wavelengths o f ‘Green’ and ‘Red’ filters the emission efficiency o f the LED lamp is nearly constant for all three settings; ii) for ‘Blue filter wavelengths the efficiency increases with lamp current and for setting 3 is even less than for setting 2. The most possible explanation of the last effect is a distortion of the lamp spectrum with current: it is warming at low current. Taking into account a wing of typical very sharp peak at the LED spectrum in the region of 427.8 nm we conclude that the light source cannot be used directly for calibration in this wavelength region. The measured emission efficiencies were normalized by measurements provided in Kiruna on 15 February 2012 and the resulted absolutely calibrated values are presented in Table 2. Table 2. Efficiency of the SID 105 emission deduced from measurements in Kiruna (15 February 2012). SID 105 setting £ 557.3 [R т Л ''] £629.9 [R mA'1] £ 428.0 [R m A 1] 1 33±3 60±13 11.5±0.4 2 58±10 89±16 15.8±0.6 3 67±17 113±45 15.4±0.6 After the study described above we decide to assemble a special stabilized power source with continuous indications of voltage and electric current. We hope this power supply in the future will always be used with the SID 105. The current-voltage characteristics of the SID 105 for intensity settings 1, 2 and 3 measured by this power supply are shown in Fig.3. The error bars are 1% for the voltage and 2% for the current. The experimental points can be inter- and extrapolated by the following empirical dependences: /set i = 50 Vsa , - 9 0 (4) Ле( 2= 300 (К** 2 - 2.95)'/г+ 19 (5) Aetj= 560 (Fset 3 - 3)'/!+ 13 (6) To apply the results of the light source calibrations to the camera calibration curves we need to deduce the electric circuit parameters used in both cases, and to find a way to renormalized light intensity of the lamp from one case to another. Taking into account Eq.3, the calibration factor can be found as [RsCTS1 - - T f > m Here for given wavelength X: ex is calibrated efficiency of the light source (Table 2), / obs is the DC current during camera calibration, Вл and кл - are the camera-dependent constants (Table 1), g is camera gain level. Comparing with Eq.2, the calibration constant is: Fx [R s CTS *] = sд/obs # л 1 (8) The results for three wavelengths are shown in Table 3. For >i=427.0 the calibration factor is increased by factor 1.6 due to trend in the efficiency of the SID 105 emission with electric current (not discussed here in details). To extrapolate the calibration to other wavelengths we can use known relative spectral response r(A), see Fig. 1. Eq.(3). Filter G-l G-2 ‘Red’ 630.0 nm *бзо.о= 0.0037 #бзо.о==0.133 *взо.о= 0.0038 # 630.0 = 0.123 ‘Green’ 557.8 nm * 557.8 ~ 0.00406 # 557.8 = 1-67 *557.8= 0.00407 #557.8= 1-63 ‘Blue’ 428.0 nm * 428.0 = 0.00405 #428.0= 1-12 * 428 . 0 “ 0.00407 #428.0= 1-08 153