The measurements of the angular dependence were performed under the solar simulator, on a cooled rotating measurement block. The change in current was determined for the incidence angle varying between -75° and +75°. The main bus bars were perpendicular to the rotation axes during the measurement. The values for positive and negative angles were always averaged. In Fig. 3a, the angular dependency of the normalised short-circuit current /sc(e)/[/sc(0) cos(e)] is plotted as a function of the incidence angle в. The curves are obtained by averaging the results obtained on 6 modules with AR layer and 6 without and show the departure from the ideal cosine law at angles higher than 45°. A significant improvement is given at high incidence angle by the AR layer, where the angular losses remain minimum. In Fig. 3b, the total current gain in % given by the AR layer is shown.

The average gain can be fitted with a single exponential growth curve. Starting from 2.65% at 0° it reaches 3% at 45°, 6% at 60° and 12% at 75°.

Fig. 3. a) Normalised current as a function of the light incidence angle for the glass without and with AR layer (The data points at 80° and 90° have been extrapolated). b) Gain in current given by the AR layer as a function of the incidence angle.

3.4 Outdoor modules

The nominal operating cell temperature (NOCT) of the modules has been determined by analysing 200 values of the module temperatures, for illuminations between 700-900 W/m2 and air temperatures between 20°-25°C. In average, the module with the AR layer was found to be 1.2°C warmer than the module with the normal glass, with an average backskin temperature of 43.4° and 42.2°C respectively (temperature at the backside of the module measured on the Tedlar foil).